The thesis by Patrick Ryan Lowenthal explores the concept of social presence in online learning, examining how it manifests in fully asynchronous courses. It critiques previous research for primarily focusing on student perceptions rather than the establishment of social presence, employing a mixed-methods approach to analyze communication in online discussions. The study finds that social presence is influenced by various situational factors such as group size and instructional tasks, suggesting complexities in fostering social presence in online education.

1. SOCIAL PRESENCE: WHAT IS IT? HOW DO WE MEASURE IT?
by
Patrick Ryan Lowenthal
B.A., Georgia State University, 1997
M.A., University of Colorado Boulder, 1999
M.A., University of Colorado Denver, 2003
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado Denver in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
Educational Leadership and Innovation
2012

2. UMI Number: 3506428
All rights reserved
INFORMATION TO ALL USERS
The quality of this reproduction is dependent on the quality of the copy submitted.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
UMI 3506428
Copyright 2012 by ProQuest LLC.
All rights reserved. This edition of the work is protected against
unauthorized copying under Title 17, United States Code.
ProQuest LLC.
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, MI 48106 - 1346

3. ii
This thesis for the Doctor of Philosophy degree by
Patrick Ryan Lowenthal
has been approved for the
Educational Leadership and Innovation
by
Joanna C. Dunlap, Chair
Joanna C. Dunlap, Advisor
Rodney Muth
Ellen Stevens
Patti Shank
Date

4. iii
Lowenthal, Patrick Ryan (Ph.D., Educational Leadership and Innovation)
Social Presence: What is it? How do we measure it?
Thesis directed by Associate Professor Joanna C. Dunlap
Social presence theory is a central concept in online learning. Hundreds of studies have
investigated social presence and online learning. However, despite the continued interest
in social presence and online learning, many questions remain about the nature and
development of social presence. Part of this might be due to the fact that the majority of
past research has focused on students' perceptions of social presence rather than on how
students actually establish their social presence in online learning environments. Using
the Community of Inquiry Framework, this study explores how social presence manifests
in a fully asynchronous online course in order to help instructional designers and faculty
understand how to intentionally design opportunities for students to establish and
maintain their social presence. This study employs a mixed-methods approach using
word count, content analysis, and constant-comparison analysis to examine threaded
discussions in a totally online graduate education course. The results of this study suggest
that social presence is more complicated than previously imagined and that situational
variables such as group size, instructional task, and previous relationships might
influence how social presence is established and maintained in threaded discussions in a
fully online course.
The form and content of this abstract are approved. I recommend its publication.
Approved: Joanna C. Dunlap

5. iv
DEDICATION
I dedicate this thesis to the ladies of my life. First, I dedicate this to my mother. I
would not be the person I am today if it was not for her. Second, I dedicate this to my
wife, Alison, for (among other things) her unfaltering support and patience while I was
avoiding completing this thesis. I could not have completed this without her love and
support. Third, I dedicate this to my daughters, Jordan and Ashlyn. I hope they
understand one day why Daddy spent so much time on the computer. And over time I
hope they see me spend less time on the computer and more time with them. Last but not
least, I dedicate this to the two greatest dogs in the world, Beezer and Nikita. They both
supported me in their own way throughout this process over the years, and I miss them
dearly now that they are gone.
.

6. v
ACKNOWLEDGEMENT
I want to thank my advisor, Joanna C. Dunlap, for her guidance, support, and
patience over the years. Joni taught me how to be a scholar and has been a great
colleague and friend. I look forward to continuing our relationship for years to come. I
also want to thank Ellen Stevens for never giving up on me and always asking those
tough questions over the years. I want to thank Rodney Muth for his unending support. I
took my first EDLI course with Rod, I published my first article with Rod, and I finished
my dissertation with Rod. I would also like to thank Marcia Muth for teaching me to be a
writer when that was the last thing I thought I would ever become. And finally I would
like to thank Patti Shank for her continued professional support over the years.

7. vi
TABLE OF CONTENTS
FIGURES ............................................................................................................... xi
TABLES .............................................................................................................. xiii
CHAPTER
1. INTRODUCTION ...............................................................................................1
Background ..................................................................................................3
Social Presence Theory ....................................................................3
The Evolution of Social Presence Theory .......................................5
Limitation of Previous Studies.....................................................................6
Statement of the Problem .............................................................................9
Conceptual Framework ..............................................................................10
Goal of the Study .......................................................................................14
Overview of Methods ................................................................................16
Sample ............................................................................................16
Data Analysis .................................................................................16
Reliability and Validity ..................................................................18
Significance of Study .................................................................................18
Limitations .................................................................................................19
Chapter Summary ......................................................................................19
2. LITERATURE REVIEW ..................................................................................21

8. A Brief History of Social Presence Theory ...............................................21
Theoretical Foundations of Social Presence Theory .....................21
Intimacy .............................................................................22
Immediacy ..........................................................................22
Influential and Related Research on Social Presence ....................23
Competing Theories of Social Presence Theory ........................................26
Cuelessness ....................................................................................26
Media Richness ..............................................................................27
Social Information Processing .......................................................28
Defining Social Presence ...........................................................................31
Measuring Social Presence ........................................................................33
Gunawardena’s Social Presence Scale ...........................................34
Rourke et al.’s Social Presence Indicators .....................................35
Tu and The Social Presence and Privacy Questionnaire ...............37
Research on Social Presence ......................................................................40
Social Presence and Student Satisfaction ......................................40
Social Presence and Interaction .....................................................44
Social Presence and Student Learning ...........................................47
Establishing and Maintaining Social Presence ..........................................53
Some Gaps in the Literature ......................................................................57
Chapter Summary ......................................................................................60
vii

9. 3. METHOD .........................................................................................................61
Research Question .....................................................................................61
Research Design .........................................................................................61
Sample ........................................................................................................62
Sampling Scheme ...........................................................................62
Sampling Design ............................................................................65
Data Collection ..........................................................................................67
Data Analysis .............................................................................................67
Word Count ....................................................................................68
Content Analysis ............................................................................69
Constant Comparison Analysis ......................................................77
Reliability and Validity ..............................................................................79
Reliability .......................................................................................79
Validity ..........................................................................................80
Chapter Summary ......................................................................................81
4. RESULTS .........................................................................................................82
Word Count ................................................................................................82
Content Analysis ........................................................................................87
Stage One: Social Presence Categories and Indicators Across All
Threaded Discussions ....................................................................89
viii

10. Stage Two: Social Presence Categories and Indicators by Threaded
Discussion ......................................................................................94
ix
Stage Three: Social Presence Categories and Indicators by
Students ........................................................................................101
Constant Comparison Analysis ................................................................106
Chapter Summary ....................................................................................111
5. DISCUSSION ..................................................................................................112
Key Findings ............................................................................................112
Group Size ...................................................................................114
Instructional Task .........................................................................116
Past Relationships ........................................................................119
One Size Does Not Fit All ...........................................................120
Limitations of Studying Social Presence .................................................121
Situational Variables of CMC ......................................................122
Unit of Analysis ...........................................................................126
Problems with the Social Presence Indicators and Treating Them
Equally .........................................................................................128
Problems with Measuring the Community of Inquiry .................130
Limitations of the Study...........................................................................132
Concluding Thoughts and Implications ...................................................133

11. x
APPENDIX
A. APPENDIX A .................................................................................................136
B. APPENDIX B .................................................................................................142
C. APPENDIX C .................................................................................................146
REFERENCES ................................................................................................................150

12. xi
LIST OF FIGURES
Figure
1.1 Community of Inquiry Framework ........................................................................11
1.2 Visual Depiction of Initial Conceptual Framework of Social Presence Developed
by Rourke et al., 2001a ..........................................................................................14
2.1 Communication Media and Information Richness Diagram .................................28
2.2 Timeline of Competing Theories of Social Presence Preceding the Development
of the Community of Inquiry Framework ..............................................................30
2.3 Continuum of Definitions of Social Presence ........................................................33
3.1 Steps Followed to Complete Constant Comparison Analysis of Online
Discussions ............................................................................................................78
4.1 Word Cloud of Word Count Results Without the Discussions Headings .............84
4.2 Frequency of Possible Social Presence Indicators Across the Three Major
and Most Frequented Threaded Discussions .........................................................85
4.3 Stages of Disaggregation of Content Analysis Used to Explore Use of Social
Presence Indicators in a Fully Online Asynchronous Course ................................88
4.4 A Visual Depiction of the Frequency of Each of the Three Social Presence
Categories ..............................................................................................................90
4.5 Frequency of Social Presence Indicators Across All Threaded
Discussions ............................................................................................................92
4.6 Social Presence Indicators Separated by Category ................................................93
4.7 Visual Depiction of the Average Social Presence Indicators Group by Category in
Closed Threaded Discussions ................................................................................97

13. xii
4.8 Ranking of Social Presence Indicators Used By the Three Students with the
Highest Overall Social Presence Per Post Average .............................................104
4.9 Disaggregation of Three Students with Highest Social Presence per Post
Average ................................................................................................................106

14. xiii
LIST OF TABLES
Table
1.1 Categories and indicators of social presence ...............................................................12
1.2 Alignment of research questions to data analysis ........................................................18
2.1 Phases of social presence research ...............................................................................30
2.2 Example of social presence indicators ...................................................................36
2.3 Social presence dimension of the Community of Inquiry Questionnaire ..............39
2.4 Strategies to establish and develop social presence ...............................................53
2.5 Strategies to establish and maintain social presence ..............................................55
3.1 Online descriptions ................................................................................................64
3.2 Threaded discussions raw data ...............................................................................66
3.3 Overview of data analysis ......................................................................................68
3.4 Original social presence categories and example indicators ..................................70
3.5 Rourke et al.’s categories and indicators of social presence ..................................71
3.6 Evolution of the indicators of social presence .......................................................72
3.7 Swan and Hughes et al. combined list of categories and indicators of
social presence .......................................................................................................73
3.8 Coding sheet used for content analysis ..................................................................75
4.1 Top 20 words used across all threaded discussions ...............................................83
4.2 Top 20 words across project groups ......................................................................86
4.3 Top 20 words across pairs ......................................................................................86
4.4 Top 20 words across reading groups .....................................................................87
4.5 Social presence frequency across all forums .........................................................91

15. 4.6 Social presence indicators ranking from highest to lowest frequency ...................92
4.7 Open vs. closed threaded discussions ....................................................................95
xiv
4.8 Average social presence indicators per post across open and closed threaded
discussions .............................................................................................................96
4.9 Average social presence indicators across closed threaded discussions ................97
4.10 Ranking of average social presence indicators across closed threaded discussions98
4.11 Average social presence indicator per threaded discussion .................................100
4.12 Student’s use of social presence categories .........................................................102
4.13 Groups of codes resulting from the constant comparison analysis of reading
Group E ................................................................................................................108
4.14 Groups of codes resulting from the constant comparison analysis of
Pair 9 ....................................................................................................................110
5.1 Teaching presence categories and indicators .......................................................113
5.2 Instructor vs. student postings in small discussions.............................................117
5.3 Measuring social presence in a Community of Inquiry .......................................131

16. 1
CHAPTER 1
INTRODUCTION
I can remember when I started teaching online. I was a full believer in online
education. I had been teaching face-to-face courses and even taken a few courses online
myself. I was excited to teach online. At the same time, I was scared. I was scared that
somehow my personality, my classroom presence, my empathy, my ability to connect
with my students—all things that I attributed to my success teaching face-to-face—would
not translate to an online environment.
I regularly meet faculty now who have similar fears. They fear that what they do
in the classroom cannot translate to an online environment. Fears like these, though, are
not restricted to faculty. I meet people all the time who make claims like, “I just can’t
learn that way” or “I need to talk to people face-to-face” or “online learning is just not for
me.” For some time, people have had the choice to avoid learning online if it was not
their preferred way to learn. But the growth of online education (see Allen & Seaman,
2006, 2010), legislative trends that require students to learn online (Walters, 2011;
Watson, 2006), and the blurring of boundaries between fully online and traditional face-to-
face courses (Woo, McNeill, Preston, Green, & Phillips, 2008), suggest that in the near
future faculty and students will no longer have the choice to avoid online education.
Based on my research and experience, I contend that one’s success learning
online—specifically in formal online education settings—begins and ends with one’s
ability to communicate effectively online. In my experience, students who struggle
communicating online (whether within a Learning Management System or using email)
struggle learning online in formal online educational settings. Communicating online is

17. 2
simply different from communicating face-to-face (Suler, 2004). I am interested in these
differences and how people—specifically faculty and students—take advantage of these
differences in formal education settings. In other words, I am interested in how faculty
and students leverage the strengths and minimize the limitations of a computer-mediated
communication (CMC) medium when teaching and learning online.
A supposed limitation of CMC and online education in general is that it is
difficult to establish one’s presence as a “real” person and “connect” with others—
generally called social presence (Kear, 2010). One reason people struggle learning online,
I posit, is related to this concept of social presence or the lack there of. For instance,
isolation and loneliness—which are in part due to a lack of presence—are often cited as
reasons why students do not persist online (Ali & Leeds, 2010; Ludwig-Hardman &
Dunlap, 2003).
I have set forth to investigate the big question of how people establish their
presence online by examining how people present themselves as real people in formal
online education environments (which predominantly rely on asynchronous CMC).
Ultimately, my hope is that my research can help others learn how to establish their social
presence in formal online education environments. In the following pages of this chapter,
I provide a formal rationale for and overview of this study by beginning with some
background literature on social presence, addressing limitations of previous research,
presenting my conceptual framework, and finally providing an overview of the
methodology used for this study.

18. 3
Background
In the late 1980s and early 1990s, researchers began to study the effects of
computer-mediated communication (CMC) (Daft & Lengel, 1984, 1986; Rutter, 1984,
1987; Walther, 1996). Some concluded that CMC was inherently antisocial and
impersonal (Walther, 1996; Walther, Anderson, & Park, 1994). While Hiltz and Turoff
(1993), two early key researchers of CMC, acknowledged that interpersonal relationships
might be fostered through CMC, early research suggested—and convinced others—that
CMC was better at task-oriented communication than interpersonal communication
(Walther & Parks, 2002). To make sense of findings like these, CMC researchers turned
to theories like Cuelessness Theory (Rutter, 1984, 1987), Media Richness Theory (Daft
& Lengel, 1984, 1986; Daft, Lengel, & Trevino, 1987), Social Information Processing
Theory (Walther, 1996; Walther & Parks, 2002) and Social Presence Theory (Short,
Williams, & Christie, 1976). Overtime, social presence theory appealed to more
researchers of online learning (as is evidenced in the growing body of research on social
presence and online learning). And today, social presence theory is the most often
referenced theory explaining the social nature of CMC in online educational
environments (Lowenthal, 2010).
Social Presence Theory
Short, Williams, and Christie (1976) originally developed the theory of social
presence to explain the effect telecommunications media have on communication. They
defined social presence as the degree of salience (i.e., quality or state of being there)
between two communicators using a communication medium. They posited that
communication media differ in their degree of social presence and that these differences

19. 4
play an important role in how people interact. They conceptualized social presence
primarily as a quality of a communication medium that can determine the way that people
interact and communicate. From their perspective, people perceive some media as having
a higher degree of social presence (e.g., video) and other media as having a lower degree
of social presence (e.g., audio) and still other media having even a lower degree of social
presence (e.g., text). More importantly, Short et al. believed that a medium with a high
degree of social presence is seen as being sociable, warm, and personal, whereas a
medium with a low degree of social presence is seen as less personal. While people might
want a less intimate or immediate communication medium from time to time (see
Williams, 1975), formal education is a very social process that involves high
interpersonal involvement. Past research, for example, has specifically stressed the
importance of contact and cooperation between faculty and students (Chickering &
Gamson, 1987). Thus, early on social presence theory appeared to have direct
implications for educators in online environments.
In the late eighties and early nineties, relying on this theory, researchers began
concluding that CMC was inherently impersonal because the nonverbal and relational
cues (common in face-to-face communication) are filtered out of CMC (Walther & Parks,
2002). Later though in the mid-nineties, researchers began to notice, even though CMC
lacks nonverbal and relational cues, that it can still be very social and interpersonal
(Gunawardena, 1995; Gunawardena & Zittle, 1997) and at times even hyperpersonal
(Walther, 1996). Further, as researchers (Gunawardena, 1995; Tu, 2000) began
examining the sociability of online education, they started questioning the degree to
which the attributes of a communication medium—in this case the cues filtered out of

20. 5
CMC systems—determine how people socially interact (Danchak, Walther, & Swan,
2001; Gunawardena, 1995; Gunawardena & Zittle, 1997; Richardson & Swan, 2003; Tu,
2000).
The Evolution of Social Presence Theory
Researchers of online learning (e.g., Gunawardena, 1995; Gunawardena & Zittle,
1997; Tu, 2000) began questioning the theory of social presence developed by Short et al.
(1976). These researchers argued, based on their experience and research, that
participants in online asynchronous discussions, using text alone, are able to project their
personalities into online discussions and create social presence. They found that online
learners are able to present themselves as being “real” as well as “connect” with others
when communicating in online learning environments by doing such things as using
emoticons, telling stories, and even using humor (Rourke et al., 2001a; Swan, 2003).
Thus, a user’s personal perceptions of social presence—which are influenced over time
and with experience using a communication medium—and the behaviors one learns to
use to make up for the cues that are filtered out matter just as much, if not more, than a
medium’s supposed capabilities. This new line of research sparked a renewed interest in
the sociability of online learning, social presence, and CMC as evidenced in the increased
amount of literature focused on social presence.
Given the research stream, social presence is now a central concept in online
learning. For instance, social presence has been listed as a key component in theoretical
frameworks for distance education (Akyol & Garrison, 2009; Benbunan-Fich, Hiltz, &
Harasim, 2005; Vrasidas & Glass, 2002). Researchers have shown—to varying degrees—
a relationship between social presence and student satisfaction (Gunawardena, 1995;

21. Gunawardena & Zittle, 1997; Hostetter & Busch, 2006; Richardson & Swan, 2003; So &
6
Brush, 2008), social presence and the development of a community of learners (Rourke,
Anderson, Garrison, & Archer, 2001a; Rovai, 2002; Ryman, Hardham, Richardson, &
Ross, 2009), and social presence and perceived learning (Caspi & Blau, 2008;
Richardson & Swan, 2003). Just as earlier researchers of CMC (Kiesler, 1986; Kiesler,
Siegel, McGuire, 1984) used social presence theory to explain why CMC was inherently
impersonal, later researchers (Gunawardena, 1995; Tu, 2000) reconceptualized social
presence theory—focusing less on the medium and more on how people adapted to the
medium—to explain how CMC in online learning environments can be very personal and
social.
Limitations of Previous Studies
Despite the intuitive appeal and overall popularity of social presence theory,
research on social presence still suffers from a few problems. Early studies of social
presence and CMC had contradictory findings (see Walther et al., 1994). For instance,
studies conducted in laboratory settings tend to support cues-filtered-out perspectives that
suggested that CMC was inherently anti-social (Connolly, Jessup, & Valacich, 1990;
Hiemstra, 1982), whereas studies conducted in the field often did not (Walther, 1992;
Walther et al., 1994; Weedman, 1991). Walther et al. (1994) explain that contradictory
findings like these are likely due to the abbreviated time periods and unrealistic
experimental settings researchers used to study CMC.
In much the same way, later research on the sociability of online learning, social
presence, and CMC suffers from a number of limitations. First, researchers of social
presence cannot agree upon a single definition of social presence (Biocca & Harms,

22. 7
2002; Biocca, Harms, & Burgoon, 2003; Rettie, 2003; Lane, 2011; Tu, 2002b). Instead,
researchers continue to redefine social presence (Lowenthal, 2010; Picciano, 2002).
Second, the majority of research conducted on social presence has various
conceptual or methodological limitations. For example, Gunawardena (1995;
Gunawardena & Zittle, 1997), one of the foundational and most often cited researchers
on social presence, primarily investigated learners’ feelings toward CMC as a medium of
communication (e.g., asking students the degree to the which they agree to statements
like “CMC is an excellent medium for social interaction”) rather than specifically asking
about how people adapted the medium for social purposes. Other researchers studied
social presence in hybrid courses (e.g., Hughes et al., 2007; Shea & Bidjerano, 2010; So
& Brush, 2008), online courses that had face-to-face meetings at the beginning of the
course (e.g., Tu, 2001; Wise et al., 2004), or non-traditional learning environments (e.g.,
6-week. self-paced, faculty-directed courses consisting of a single student) (e.g., Wise,
Chang, Duffy, & Del Valle, 2004). Each of these contexts would inevitably influence
how one establishes his or her own social presence as well as how one perceived the
social presence of others, but researchers (e.g., Richardson & Swan, 2003; Swan & Shih,
2005) have not explicitly acknowledged how these differences influence social presence.
In addition, most researchers studying social presence (e.g., Arbaugh &
Benbunan-fich, 2006; Garrison, Cleveland-Innes, & Fung, 2010; Gunawardena, 1995; Tu
2002a; Richardson & Swan, 2003) have used similar data-analysis techniques. The
majority of research has relied either on content analysis or on self-report data (obtained
through a questionnaire). Relying solely on one type of analysis can lead researchers to
make interpretive errors about the underlying phenomenon they are studying (Leech &

23. 8
Onwuegbuzie, 2007). Studies of social presence might benefit from employing multiple
or mixed methods (see Lowenthal & Leech, 2009).
Third, foundational research on social presence is dated (Gunawardena, 1995;
Gunawardena & Zittle, 1997; Rourke et al, 2001a; Tu, 2001, 2002a, 2002b). The
majority of the foundational research on social presence is over five to ten years old, and
during the past five years alone CMC and online learning have grown exponentially.
CMC is no longer a fringe activity used by a select group of users (Smith, 2010); rather,
CMC, issues of the digital divide aside, is commonplace. As people use the Internet and
email to communicate with others more each day, it is logical to assume that they become
more adept at communicating, becoming literate with this medium. This is not simply a
case of supposed “digital natives” (i.e., those who have grown up with technology) using
CMC differently than “digital immigrants” (i.e., those who are new to technology)
(Brown, 2002; Prensky, 2001). Rather, it is an issue of how people learn to use any
communication medium better over time: The cell phone is a perfect example with
millions of users worldwide, from the slums of India to the penthouses of New York
City—nearly everybody seems to have a cell phone these days.
The increased amount of time spent online has led online users of all ages and all
generations to adjust their perceptions, expectations, and day-to-day use of CMC. Just as
research in the early 1990s (e.g., Gunawardena, 1995; Walther, 1992, 1994, 1996) began
to call into question CMC research in the 1980s (i.e., Kiesler, 1986; Kiesler, Siegel,
McGuire, 1984; Rutter, 1984, 1987), additional research on social presence might begin
to question research conducted in the late 1990s and early 2000s (e.g., Gunawardena &
Zittle, 1997; Rourke et al, 2001a; Tu, 2000). Researchers need to continue to study social

24. 9
presence, and at times even replicate previous studies (unfortunately rarely done), in
order to ensure that current assumptions about social presence are still correct across
various contexts.
Finally, and most important, some research on social presence contradicts other
research (see Lowenthal, 2010). For instance, some researchers have found that social-presence
behaviors used by online learners decrease over time (Rourke, Anderson,
Garrison, & Archer, 2001a), while others have found that social presence behaviors do
not decrease over time (Stacey, 2002). In addition, Picciano (2002) found a relationship
between social presence and student learning, while Wise et al. (2004) did not. For all of
these reasons, additional research on social presence in online learning environments is
needed—and especially in asynchronous learning environments, the dominant form of
online education (National Center for Education Statistics, 2008)—to help clarify what
social presence is and its role in online learning.
Statement of the Problem
Despite the continued interest in social presence and CMC, many questions
remain about the nature and development of social presence (Lowenthal & Dunlap, 2011;
Swan & Shih, 2005; Rourke & Kanuka, 2009). In addition, some of what researchers and
practitioners think they do know is questionable due to the limitations of past research.
The majority of research on social presence (e.g., Gunawardena, 1995; Na Ubon &
Kimble, 2003; Picciano, 2002; Richardson & Swan, 2003; Rourke & Anderson, 2002b;
Russo & Campbell, 2004; Tu, 2002b; Wheeler, 2005; So & Brush, 2008) has focused on
faculty and students perceptions of social presence. Fewer studies by comparison (e.g.,
Hughes, Ventura, & Dando, 2007; Lomicka & Lord, 2007; Rourke et al., 2001a; Swan,

25. 10
2002, 2003a) have actually studied observable indicators of social presence in online
discussions.
While it is important to understand perceptions of social presence, it is also
important to study what students do and say online (Kramer, Oh, & Fussell, 2006).
However, not enough studies do just this and the few studies that have done this have
failed to describe adequately how social presence manifests itself in asynchronous online
courses. Researchers (e.g., Hughes et al., 2007; Rourke et al., 2001a) have typically
sampled only one part of a course and analyzed it with only one type of analysis,
typically content analysis. As a result, I posit that both researchers and practitioners may
have a very limited understanding of social presence.
Given these reasons, I set forth to conduct a mixed methods exploratory study of
social presence. I chose to do this in hopes of learning more about the observable
indicators of social presence in online course discussions.
Conceptual Framework
Many researchers (Arbaugh, 2007; Delfino, & Manca, 2007; Lomicka & Lord,
2007; Nippard & Murphy, 2007; Rourke & Anderson, 2002a, 2002b; Swan et al., 2008)
have argued for some time that the community of inquiry (CoI) framework is the most
popular framework to study social presence. The CoI framework is a comprehensive
guide (Garrison, Anderson, & Archer, 2000) for research on the practice of online
learning (Garrison & Arbaugh, 2007). Garrison et al. (2000) argued that meaningful
learning takes place in a CoI, comprised of teachers and students, through the interaction
of three core elements: cognitive presence, social presence, and teaching presence (see
Figure 1.1).

26. 11
Cognitive presence, the first element in the model, is “the extent to which the
participants in. . . a community of inquiry are able to construct meaning through
sustained communication” (Garrison et al., 2000, p. 89). Social presence, the second
element in the model, is the “ability of participants in a community of inquiry to project
their personal characteristics into the community, thereby presenting themselves to other
participants as ‘real people’” (p. 89). Finally, teaching presence, the third element in the
model, is the ability of a teacher or teachers to support and enhance social and cognitive
presence through instructional management, building understanding, and direct
instruction.
Educational
Experience
Teaching
Presence
Social
Presence
Figure 1.1. Community of inquiry framework
Cognitive
Presence
Garrison et al. (2000) initially developed three categories of social presence (i.e.,
Emotional Expression, Open Communication, and Group Cohesion). They later
developed specific indicators of social presence (e.g., use of humor, continuing a thread,
or the use of vocatives) (Rourke et al., 2001a) to help identify observable instances of
social presence in CMC (see Table 1.1). They later renamed these categories (e.g.,

27. 12
Emotional Expression was renamed Affective Responses) and tested the validity of the
categories and indicators of social presence (Rourke et al., 2001a). Swan (2003)
expanded the indicators even further, and then Hughes et al. (2007) later (though
apparently unaware of Swan’s work) made some changes to Rourke et al.’s indicators as
well. Despite the renaming of the categories and some minor changes to the social
presence indicators (which are discussed in more detail in Chapters 2 and 3), Garrison et
al.’s (2000) original categories and the later complete list of indicators (Rourke et al.,
2001) of social presence have—for the most part—remained unchanged (see Table 1.1).
Table 1.1 Categories and Indicators of Social Presence
Category Indicators Definition of Indicators
Affective
Expression of
Responses
emotions
(originally
“Emotional
Expression”)
Conventional expressions of emotion, or
unconventional expressions of emotion,
includes repetitious punctuation,
conspicuous capitalization, emoticons
Use of Humor Teasing, cajoling, irony, understatements,
sarcasm
Self-Disclosure Presents details of life outside of class, or
expresses vulnerability
Interactive
Responses
(originally
“Open
Communication”)
Continuing a Thread Using reply feature of software, rather than
starting a new thread
Quoting from Other
Messages
Using software features to quote others
entire message or cutting and pasting
sections of others’ messages
Referring explicitly
to other messages
Direct references to contents of others’ posts
Asking questions Students ask questions of other students or
the moderator
Complimenting,
expressing
appreciation
Complimenting others or contents of others’
messages
Expressing
agreement
Expressing agreement with others or content
of others’ messages

28. 13
Table 1.1 (con’t.)
Cohesive
Responses
(originally
“Group
Cohesion”)
Vocatives Addressing or referring to participants by
name
Addresses or refers
to the group using
inclusive pronouns
Addresses the group as we, us, our, group
Phatics / Salutations Communication that serves a purely social
function; greetings, closures
Note. From “Assessing Social Presence in Asynchronous Text-based Computer
Conferencing,” by L. Rourke, D. R. Garrison, and W. Archer, 2001a, in Journal of
Distance Education, 14.
Garrison, though, pointed out in 2008 that these indicators have not been revisited since
their initial development and that they might need to be revised (Arbaugh et al., 2008)—
which in many ways is a possible outcome of this study.
Rourke et al. (2001a) were the first to test and validate the indicators of social
presence. However, Garrison et al. (2000) and later Rourke et al. (2001a) did not clearly
identify the relationship between the indicators of social presence. In other words, they
left researchers wondering whether certain categories or indicators of social presence are
better examples than others. When faced with the need to calculate a social presence
score—from the frequency of indicators found in the coded transcripts of CMC—they
decided to treat all indicators equally and simply sum the frequencies of all 12 indicators
(Rourke et al., 2001a). This appeared to have been more of a pragmatic decision rather
than a theoretical or empirical decision to find a way to create a social presence score
from the indicators in order to quantify and compare transcripts of CMC. Rourke et al.,
2001a though openly admitted their uncertainty about weighting all 12 indicators equally.
Despite this admitted uncertainty, researchers have followed the same process in

29. 14
developing a social presence score, though Hughes et al. (2007) was openly critical of
this practice.
Following the work of researchers like Rourke et al. (2001a), I conceptualize
social presence as an additive process in which all categories and indicators of social
presence are of equal importance (see Figure 1.2). However, like Hughes et al. (2007), I
am skeptical of this conceptualization and hope that among other things my research will
(by using multiple forms of analysis) help support or challenge the assumed additive
nature of Rourke et al.’s conceptualization of social presence.
Cohesive
• Vocatives
Interactive
• Continuing a
Thread
+ + =
Affective
Responses
• Expression
of emotions
• Use of
Humor
• Self‐
Disclosure
• Use of
Inclusive
Pronouns
• Phatics /
Salutations
• Quoting from
Other Messages
• Referring
Explicitly to
Other Messages
• Asking
Questions
• Complimenting
/ Expressing
Appreciation
Social
Presence
Figure 1.2. Visual depiction of initial conceptual framework of social presence
developed by Rourke et al., 2000a.
Goal of the Study
The goal of this study is to understand better how social presence manifests in
threaded discussions in asynchronous online courses. However, all CMC is not the same
(Herring, 2007). While researchers can generalize about CMC at some level, they should

30. 15
recognize the situated and changing nature of social presence. Given this and to
accomplish the goal of this study, I study social presence in an intentional, socially
situated, specific context. Thus, the goal of this study is to explore the phenomenon
known as social presence by investigating how it manifests during online discourse in an
asynchronous online graduate education course.
The following research question guides this exploratory study: How does social
presence manifest in an asynchronous, online graduate-education course? This specific
question was chosen because the majority of research on social presence has either relied
solely on self-report data of faculty and student perceptions of social presence or has been
confined to a monomethod approach—usually using content analysis—to analyze a few
weeks of online threaded discussions. Both of these approaches fail to explore and
describe how social presence manifests in threaded discussions over the length of a
course. In other words, what are faculty and students actually doing to establish their
social presence? The focus of this study, given this research question, is on developing a
rich description of social presence by using multiple types of data analysis in order to
help faculty and students have better experiences in online courses and to enable course
designers to develop better online courses.
Overview of Methods
In the following paragraphs, I briefly describe the methods used for this study. I
specifically focus on the sample, data analysis, reliability, and validity. Each of these
topics is addressed in greater detail in Chapter 3.

31. 16
Sample
A single, completely online graduate course in education was purposefully and
conveniently sampled for this study. Thus, a non-random (non-probability) criterion
sampling scheme was used in this study (Onwuegbuzie & Collins, 2007). A section of
EDLI 7210 Educational Policy Making in a Democratic Society—which was taught
online in the spring of 2007—was identified as an appropriate sample for this study. The
course was a graduate-level online course in the School of Education and Human
Development at the University of Colorado Denver delivered via eCollege. All of the
threaded discussions in the eCollege course shell for this course were used for this study.
The population of the course primarily consisted of graduate students completing
coursework for an Educational Specialist (EdS) degree or a PhD. Many of the EdS
students were also seeking their principal license. Nineteen graduate students were
enrolled in the course.
Data Analysis
The majority of research on social presence has relied primarily on self-report
survey data (e.g., Gunawardena, 1995; Richardson & Swan, 2003). While self-report
survey measures are useful and have their place in educational research, as Kramer, Oh,
and Fussell (2006) point out, they “are retroactive and insensitive to changes in presence
over the course of an interaction [or semester]” (p. 1). In this study, rather than focus on
students’ perceptions of presence (which I have done in other studies such as Lowenthal
& Dunlap, 2011; Lowenthal, Lowenthal, & White, 2009), I focused instead on what was
“said” in the online threaded discussions.

32. 17
I used a mixed-methods exploratory methodology (Miles & Huberman, 1994;
Onwuegbuzie & Leech, 2005b) that employed both quantitative and qualitative methods
to conduct this study. In order to explore social presence in a specific situated
asynchronous learning environment in great detail, I analyzed the online threaded
discussions (now archived in the discussion forums) using word count, content analysis,
and constant comparison analysis (Leech & Onwuegbuzie, 2007).
More specifically, multiple forms of data analysis were used to address the
research question— How does social presence manifest in a graduate education
asynchronous online course? (see Table 1.2 above for an illustration of this). First, I
analyzed all of the discussions with word count (in conjunction with basic descriptive
statistics of each forum) to identify which threaded discussion had a higher frequency of
words and posts as well as which one’s had a higher number of social presence indicators
(types of words). Second, I used content analysis to analyze every threaded discussion,
using a modified version of the social presence indicators developed by Garrison et al.
(2000) and later modified by Swan (2003) and Hughes et al. (2007). Based on the results
of the word count and content analysis, I then selected two discussion threads—one with
a high number of social presence indicators and one with a low number of social presence
indicators—to analyze in more depth with a grounded theory constant comparison
analysis technique.

33. 18
Table 1.2 Alignment of Research Questions to Data Analysis
Research Question Data Analysis Type of Data
How does social presence
manifest in a graduate
education asynchronous
online course?
• Word Count
(Quantitative)
• Content Analysis
(Quantitative)
• Constant Comparative
Analysis
(Qualitative)
• All course discussions
• All course discussions
• One discussion threads
with high social
presence & one with
low social presence
Reliability and Validity
Reliability and validity are key considerations for any researcher. The most
common method used to calculate interrater reliability is a percent agreement statistic
(Rourke et al., 2001b). Two researchers (me and another researcher) coded the threaded
discussions using content analysis. A percent agreement statistic was calculated using
Holsti’s (1969) coefficient of reliability. A large component of establishing validity—
which is often described as trustworthiness in qualitative literature—is developing a
sound theoretical framework (Garrison, Cleveland-Innes, Koole, & Kappelman, 2006). I
have established the validity of this study by working from Garrison et al.’s CoI
framework. Further, the coding schemes I used for this study also came directly from the
literature (Hughes et al., 2007; Rourke et al., 2001a; Swan, 2003).
Significance of the Study
Learning is a very human and social activity (Dunlap & Lowenthal, 2009b).
Online learning environments, though, can feel isolating and impersonal. Given this,
educators must find ways to make formal online learning environments more personal

34. 19
and less isolating not only to help students persist but also to increase engagement and
satisfaction. To accomplish this, educators have focused on establishing social presence
in online courses (Dunlap & Lowenthal, 2009b).
The significance or educational value of this research lies in its ability to help
researchers better identify and study instances of social presence as well as to help faculty
who teach online better understand how they can identify and establish social presence by
using specific indicators of social presence. Further, the results of this study can help
instructional designers design and develop online courses that utilize specific
instructional approaches to help students establish their social presence online.
Limitations
All studies suffer from some type of limitation. Perhaps the most obvious
limitation is the time that has passed between when the course was offered and when I
analyzed the data. Related to this limitation is my inability to check with students
(whether through specific interviews or member checking) to verify whether or not what
I found in the course discussions is actually what they intended. However, one of the
main reasons to focus on the language students use is because students rarely clarify what
they mean by a posting; rather, other students simply do their best to make sense of what
they read. In other words, in my experience very little member checking occurs in a
typical online discussion so this limitation might actually end up being a very realistic
component to this study.
Chapter Summary
Researchers have been studying social presence in online learning environments
for a number of years now (Lowenthal, 2009). However, research on social presence to

35. 20
date suffers from a host of problems—ranging from inconsistent and contradictory
findings to strange sampling decisions. Part of the problem might be the methodological
decisions made by researchers. Instead of using a monomethod approach like the
majority of past research, I employed a mixed-methods approach to studying social
presence, utilizing both quantitative and qualitative methods to investigate the complex
nature of social presence. In addition, this study specifically focused on how social
presence manifests during threaded discussions in asynchronous online courses.
In Chapter 2, I present a review of the literature. In Chapter 3, I go over the
methods used for this study. In Chapters 4 and 5 I present the results, discuss the
findings, and provide recommendations for faculty and instructional designers as well as
for future research on social presence.

36. 21
CHAPTER 2
LITERATURE REVIEW
In the following chapter, I synthesize past research on social presence in general
and specifically research on the community of inquiry (CoI) framework to provide a
foundation and some background for my study. I begin by addressing the history of
social presence theory. After that, I address some early competing theories of social
presence and some differences in how researchers define and measure social presence. I
then conclude this chapter by synthesizing some of the research conducted on the
community of inquiry in general and social presence in particular and addressing some
gaps in the literature.
A Brief History of Social Presence Theory
As mentioned in Chapter 1, Short, Williams, and Christie (1976) developed the
initial theory of social presence in their book, The Social Psychology of
Telecommunications. While this book often serves as the foundational text to understand
the initial theory of social presence, it is important to look at the foundations of this
theory as well as later research conducted by Short et al. to understand how the theory of
social presence has evolved over the years.
Theoretical Foundations of Social Presence Theory
The collective work of Short et al. (1976) that is presented in The Social
Psychology of Telecommunications as well the work Short, Williams, and Christie (e.g.,
Short, 1974; Christie & Kingan, 1977; Williams, 1975; Wilson & Williams, 1977)
conducted individually or with other colleagues before and after their seminal text was
influenced by the social psychology concepts of intimacy and immediacy. Short et al.

37. openly acknowledge that their concept of social presence is related to these two concepts.
22
Thus, each of these concepts is discussed in more detail in the following paragraphs.
Intimacy. Argyle and Dean (1965) were the first to use the concept of intimacy to
explain communication behavior. They developed a theory of intimacy and equilibrium
to explain how people communicating with each other will adjust their behavior to
maintain a sense of equilibrium. They explain that
aspects of intimacy are governed by both approach and avoidance forces, and are
kept in a condition of equilibrium for any two people…if this equilibrium is
disturbed along one of its constituent dimensions, e.g., by increasing physical
proximity, there will be compensatory changes along the other dimensions. (p.
304)
According to Argle (1969), people establish intimacy in a number of ways when
communicating, such as proximity, eye contact, smiling, and personal topics of
conversation. Short et al. (1976) argue that the social presence of a communication
medium also effects intimacy and therefore should be added to this list of ways that
people establish intimacy.
Immediacy. Wiener and Mehrabian (1968) developed the concept of immediacy.
They conceptualized immediacy as the psychological distance people put between
themselves and others when communicating. While Wiener and Mehrabian (1968) were
initially focused on speech communication, Mehrabian (1972) later distinguished
between three types of immediacy: verbal, nonverbal, and technological immediacy.
Verbal immediacy describes how people use their choice of words to reduce or increase
psychological distance between them and others. For example, the use of the words “let
us” or “we” can create more immediacy between two people than simply using “you” or
“I.”

38. 23
People also convey immediacy nonverbally through their dress, facial
expressions, or physical proximity (Mehrabian, 1972). Finally, technological immediacy
suggests that a medium of communication can convey immediacy. According to
Mehrabian (1972), communicating face-to-face is more immediate than communicating
with video; further, communicating with a video is more immediate than communicating
by phone.
While immediacy in general, and technological immediacy in particular, is similar
to social presence, Short et al. (1976) argue that important differences exist. For instance,
Short et al. argue that “for any given medium of communication (e.g., telephone) and
situation (e.g., long-distance call), immediacy may vary even when social presence does
not” (p. 73).
While Short et al. (1976) claim that important differences are found between
immediacy and social presence, the distinction is not very clear. Further, they spend only
a few paragraphs addressing the similarities and differences between social presence,
intimacy, and immediacy. Not surprisingly, subsequent researchers often fail to
differentiate clearly between intimacy, immediacy and social presence; in fact,
researchers often appear to use the terms immediacy and social presence synonymously
(e.g., Gunawardena, 1995).
Influential and Related Research on Social Presence
Short et al. (1976) were all part of the Communications Studies Group at
University College in London. The Communications Studies Group consisted of an
estimated 30 people who conducted a number of experiments in the early 1970s on
communication media (Pye & Williams, 1978). Interestingly, The Social Psychology of

39. 24
Telecommunications appears to be the only joint publication by these three researchers.
However, each of them published, as individuals or with other colleagues, a number of
other studies on the effects of communication media (e.g., Short, 1974; Christie &
Holloway, 1975; Christie & Kingan, 1977; Williams, 1975; Williams, 1977; Wilson &
Williams, 1977). The majority of this research focused on comparing people’s attitudes
toward different communication media (e.g., face-to-face, audio, video). The following
paragraphs briefly summarize a few key findings from this early research that later
influenced the development of and people’s understandings of social presence theory.
The majority of this early research focused on the assumed importance of the
visual channel of communication. Given the importance placed on the visual channel in
previous literature, Short et al. (1976) and colleagues not surprisingly found that the
visual channel of communication was an advantage of a communication medium and
therefore highly important (Christie, 1974; Short, 1974; Williams, 1975). Christie (1974)
reports from one study that
visual media were judged more useful for complex group discussions, private
conversations and non-private dyadic conversations. Thus, the presence of visual
channel appears to be perceived as an important advantage of a communications
medium. (p. 367)
Additional research (Christie, 1974; Christie & Kingan, 1977; Williams, 1975),
though, began to show that the importance of a communication medium depended largely
on the task at hand. In fact, according to Christie (1974), “it is clearly misleading to
conceptualize different media as lying along a single dimension of acceptability or
usefulness. Their perceived usefulness varies according to the application considered” (p.
368). Williams (1975) argued that people might want a less intimate or immediate
communication medium for certain tasks. For instance, Williams (1975) suggests “that

40. 25
with tasks of very high intimacy—perhaps very embarrassing, personal or conflictual
ones—the least immediate medium, the telephone, would lead to more favorable
evaluations than either of the more immediate media” (p. 128). Further, their research
showed that tasks that are low on interpersonal involvement but still cooperative in nature
can easily be accomplished by audio or video conferencing (Williams, 1978a); however,
tasks that require more interpersonal involvement “are sensitive to the substitution of
telecommunications for face-to-face interaction” (p. 127).
Other than the suggestions made by Williams (1978a), very little was written in
these early articles about the role of the visual channel for instructional tasks. However,
Williams (1978a) argued that “tele-education seems especially promising since
educational activities are primarily for cooperative problem-solving and the transmission
of information—activities which have been shown to be almost unaffected by the
medium of communication used” (p. 129). Williams (1978a) went on to point out that our
knowledge about the role of mediated communication is far from complete—as was our
understanding of how people learned in the late 1970s.
Later research conducted by Christie and Kingan (1977), showed, among other
things, that while visual cues are helpful, they are not necessary for people to
communicate effectively. In fact, physical presence (i.e., being close to someone
physically) may be even more important for two people communicating than visual cues
(i.e., seeing another person) (Williams, 1978b). Results like these began to call for a
more complex explanation for the role of visual cues in the communication process.
Williams (1978b) suggested that the answers might be found in the theory of social
presence.

41. 26
Competing Theories of Social Presence
The theory of social presence developed by Short et al. was only one of a number
of theories used to explain the influence a communication medium can have on
communication. The three most popular competing theories of social presence—
especially during the 1980s—were Cuelessness Theory developed by Rutter (1984,
1987), Media Richness Theory developed by Daft and Lengel (1984, 1986; Daft, Lengel,
& Trevino, 1987), and Social Information Processing Theory developed by Walther
(1996; Walther & Parks, 2002). The first two theories (like Social Presence Theory) have
been described as deficit models because they focus on the cues that are filtered out and
idealize face-to-face communication as the gold standard (Thurlow, Lengel, & Tomic,
2004), whereas the third theory focuses not only on what is filtered out but what is gained
through CMC. Each of these theories are addressed briefly in the following sections to
illustrate the zeitgeist of the 1980s and early 1990s when researchers of online learning
reinvented the theory of social presence developed by Short et al.
Cuelessness
Working from a similar theoretical framework, Rutter (1984, 1987; Rutter,
Pennington, Dewey, & Swain, 1984; Kemp & Rutter, 1986) developed what he called the
Cuelessness Model. Rutter was concerned with the over emphasis placed on the
importance of eye-contact when two people were communicating. As a result, he and his
colleagues (1984) set forth to challenge the intimacy model developed by Argyle and
Dean (1965) and later Argyle and Cook (1976). Rutter and his colleagues argued that
previous research had focused too much on looking and eye-gaze and not enough on the
mutual gazing back and forth. Like Williams before, Rutter et al. (1986) found that what

42. 27
mattered was visual access to the entire person rather than simply access to another’s
eyes. They argued that it was the combined social cues—from vision and other senses—
that mattered.
The Cuelessness Model essentially claims that the fewer social cues, the greater
the psychological distance between two communicators (Rutter et al., 1986). Further, the
greater the psychological distance, the more communication turns to task-oriented
depersonalized content (Kemp & Rutter, 1986; Rutter, 1984; Rutter et al., 1986). In fact,
Rutter and colleagues (Rutter, 1989) found that the number of social cues (i.e., both
visual and physical presence cues) decreased when comparing how people communicated
in certain situations (e.g., closed-circuit television, curtain, and audio).
Media Richness
Another competing theory that emerged during the 1980s is the theory of Media
Richness. Media Richness Theory was developed by Daft and Lengel (1984, 1986).
Whereas Rutter and colleagues were aware of the work of Short et al., Daft and Lengel
never seem to explicitly acknowledge the work of Short et al. Daft and Lengel (1984)
were focused primarily on the information processing behaviors in organizations.
Therefore, they were interested in a concept called information richness:
Richness is defined as the potential information-carrying capacity of data. If the
communication of an item of data, such as a wink, provides substantial new
understanding, it would be considered rich. If the datum provides little
understanding, it would be low in richness. (p. 196)
They posited that a communication medium can determine the richness of information
(Daft & Lengel, 1986). They argued that face-to-face communication had the highest
richness and numeric communication (e.g., spreadsheet with numbers) the lowest; see
Figure 2.1 for a complete list of media richness by media.

43. 28
Information Medium Information Richness
Face-to-Face Highest
Telephone High
Written, Personal
(bulletins, documents)
Moderate
Written, Formal (bulletins,
documents)
Low
Numeric Formal
(computer output)
Lowest
Figure 2.1. Communication media and information richness diagram
Note. From “Information Richness: A New Approach to Managerial Behavior and
Organizational Design,” by R. L. Daft and R. H. Lengel, 1984, in L. L. Cummings & B.
M. Staw (Eds.), Research in Organizational Behavior (191-233). Homewood, IL: JAI.
Social Information Processing
The last of the three competing models is the Social Information Processing
model developed by Walther (1992, 1994, 1996). Walther developed his model in
response to the previous so-called “deficit” theories. Whereas previous researchers were
interested in media effects across various communication media, Walther focused
primarily on CMC. He criticized previous research, like that addressed earlier in this
chapter, for a number of reasons. First, the majority of the early research was conducted
in experimental settings that did not mirror how people communicate with different
media in real life (1992). Second, these early studies and researchers assumed that the
absence of visual cues led to an absence of sociability. Third, they assumed that task-oriented
communication lacked relational and social communication. Finally, they failed
to acknowledge that just as cues are filtered out, other cues are filtered into CMC and

44. therefore CMC has some affordances that face-to-face communication does not (Walther,
29
1996; Walther & Parks, 2002).
Walther (1992) argued that Humans’ social nature is the same in CMC and face-to-
face environments. Given enough time, he believed that people will find ways to
compensate for any cues that are filtered out in CMC. The social information processing
model essentially posits that given enough time, CMC can be very personal and even
hyperpersonal (Walther, 1992, 1996). Previous research tended to put time restrictions on
how people communicated that Walther believed diminished the possibility of
interpersonal and relational communication. Walther’s research on the other hand
suggested that
• Previous interaction between communicators influenced how people
communicated online;
• The possibility of future interaction influenced the degree to which people
socially interacted online;
• The way users used emoticons influenced interpersonal communication
online.
These competing theories help illustrate the way that thinking about a medium’s
effect on communication—especially interpersonal and social communication—change
over time. The research that began with the work of Gunawardena (1995; Gunawardena
& Zittle, 1997)—which I refer to as the third phase of social presence research (see Table
2.1 and Figure 2.2)—was influenced by previous research and theories, especially that of
Walther. Rather than conceptualizing social presence as Short et al. did, Gunawardena
and those that followed her (like Garrison et al., 2000, whose work serves as the

45. 30
conceptual framework for this study) began reconceptualizing social presence theory—
focusing more on how people appropriate technology rather than simply on what a
technology allows us to do. In fact, the work of Garrison et al. and the CoI really
represent a fourth phase of research on social presence (see Table 2.1 and Figure 2.2).
Table 2.1 Phases of Social Presence Research
Phase Period Key Figures Focus of Research
Phase 1
1970s Short et al. Focused on
Telecommunications
Phase 2 1980s-early1990s Rutter
Daft & Lengel
Kiesler
Walther
Focused on CMC
Phase 3 1990 - 1999 Gunawardena
Rourke et al.
Tu
Focused on Online
Learning
Phase 4 2000s - Present Garrison et al.
Karen Swan
Peter Shea
Focused on Social
Presence’s Role in
establishing a community
of inquiry in Online
Learning
Figure 2.2. Timeline of competing theories of social presence preceding the
development of the community of inquiry framework.

46. 31
Defining Social Presence
Given the evolution of social presence theory, it is probably not surprising that
there is not a clear, agreed upon, definition of social presence (Rettie, 2003; Tu, 2002b).
In fact, nearly everyone who writes about social presence seems to define it just a little
differently.
Presence is a key theoretical construct used in a variety of disciplines besides
communication and online learning—most notably virtual reality (see Biocca, 1997). In
fact, Lombard and Ditton (1997) identified six interrelated but distinct ways that people
understand “presence”: (a) presence as social richness, (b) presence as realism, (c)
presence as transportation, (d) presence as immersion, (e) presence as social actor within
medium, and (f) presence as medium as social actor. They even attempted to create one
all encompassing definition of presence. According to Lombard and Ditto, the following
definition takes into consideration all six ways presence is understood; presence is “the
perceptual illusion of nonmediation” (presence explicated section). To date, though, their
all encompassing definition has not been widely adopted by others. Biocca, Harms, and
Burgoon (2003) also recognized the different ways researchers across different fields
define presence. They attempted to create an all-encompassing definition of social
presence as well; they defined social presence as simply a “‘sense of being with another’”
(p. 456) whether that other is human or artificial.
Despite attempts by Lombard and Ditto (1997) and Biocca et al. (2003) to
develop some conceptual clarity when it comes to discussions of presence in general or
social presence in particular, researchers of social presence and CMC in educational
environments continue to redefine social presence (Picciano, 2002). Gunawardena (1995)

47. defined social presence as “the degree to which a person is perceived as a ‘real person’ in
32
mediated communication” (p. 151). Garrison et al. (2000), on the other hand, originally
defined social presence “as the ability of participants in a community of inquiry to project
themselves socially and emotionally, as ‘real’ people (i.e., their full personality), through
the medium of communication being used” (p. 94). Tu and McIsaac (2002) define social
presence as “the degree of feeling, perception, and reaction of being connected by CMC
to another intellectual entity through a text-based encounter” (p. 140). Finally, Picciano
(2002) defines social presence as “a student’s sense of being in and belonging in a course
and the ability to interact with other students and an instructor” (p. 22).
The differences in how researchers define social presence might seem minor but
they are important (see Ice, Gibson, Boston, & Becher, 2011). For instance, Rourke et al.
(2001) focus on students (or instructors) ability to project themselves as “real” whereas
Picciano focuses more on students’ sense of belonging to a community. Issues of
definition are important because the way researchers define social presence influences
how they measure social presence and the conclusions they draw.
Definitions of social presence, at least for researchers of social presence and
online learning, tend to fall on a continuum (see Figure 2.3). At one end of the
continuum, researchers tend to conceptualize social presence as the degree to which a
person is perceived as being “real” and being “there.” These definitions tend to focus on
whether someone is able to project himself or herself as being “real” in an online
environment and whether others perceived this person as being there and being real. In
fact, Williams (1978a) defined social presence in this way when he defined social

48. presence as “the feeling of contact obtained. . .” across various communication media (p.
33
127).
At the other end of the continuum, researchers tend to go beyond whether
someone is perceived as being “present”—that is, simply “there” or “real”—instead
focusing on whether there is an interpersonal emotional connection between
communicators. It is important to note, though, that on this end of the continuum, there
tends to be an assumption that the interpersonal and emotional connection that
communicators establish when there is social presence is a positive connection. Finally,
like most continuums, the majority of researchers find themselves somewhere in the
middle—placing some emphasis on an emotional connection—rather than on the ends of
the continuums.
Sense that someone is real
Sense that someone is there (present)
Sense that someone is real
Sense that someone is there (present)
No focus on emotion
Figure 2.3. Continuum of Definitions of Social Presence
Measuring Social Presence
Emotional Connection
After all the theorizing, researchers need to be able to identify, measure, and test
their theories about social presence. As researchers began to conceptualize social
presence differently, rather than use techniques developed and utilized by past
researchers—perhaps because of Walther’s critique of these techniques—they began to
look for new ways to study social presence. Gunawardena (1995), Rourke et al. (2001),
and Tu (2002b) have been three foundational researchers in developing ways to study

49. 34
social presence. But just like in the mid-1970s—when researchers either studied social
presence by observing user behavior or examining users attitudes (Christie, 1974)—
researchers in the third and fourth wave of social presence research have tended to focus
either on studying users’ attitudes or their behaviors online. For instance, Gunawardena
and Tu focused primarily on studying users’ attitudes whereas Rourke et al. focused on
studying users’ behaviors (though it is important to note that while Garrison early on
focused on studying users’ behaviors with his colleagues Rourke et al., he later turned to
studying students attitudes). Regardless of their focus, the work of each of these
researchers has heavily influenced most of the studies on social presence and CMC
during the past ten years. In the following paragraphs, I will address how each of these
researchers studied social presence.
Gunawardena’s Social Presence Scales
Gunawardena (1995; Gunawardena & Zittle, 1997) conducted some of the earliest
studies on social presence and CMC in an education setting. In her first article,
Gunawardena (1995) reported on two different studies she conducted in the early 1990s.
In the first study, she measured users’ perceptions of CMC using a survey. She had
students rank 17 bi-polar scales on a 5-point likert-type scale (from negative to positive).
For instance, she asked students whether CMC was more socialable or unsocialable or
more warm or cold (see Table A1 in Appendix A for the complete list). The bi-polar
scales she used focus on users’ perceptions of the medium more than the degree to which
users perceive others as “real” or “there.”

50. 35
Gunawardena (1995) reports in the same article about a second study in which she
qualitatively analyzed some data; however, she does not elaborate on what data she
analyzed or how she analyzed the data that she reported.
In a later article, Gunawardena and Zittle (1997) reported on additional data
collected from an earlier sample. However, with this study, Gunawardena and Zittle
created an instrument they called the Social Presence Scale (see Appendix A). The Social
Presence Scale was similar to the previous scale used by Gunawardena, but instead of
responding to bi-polar scales, students were asked to rank 14 questions on a scale of 1 to
5. For instance, one question asked students to rank on a scale of 1 to 5 to what degree
they agree or disagree that CMC is an excellent medium for social interaction. The Social
Presence Scale was tested for internal consistency (Alpha = .88) and appears to
investigate the construct of social presence more directly than the previous scale.
Rourke et al.’s Social Presence Indicators
Unlike Gunawardena who measured social presence through a self-report
questionnaire, Rourke et al. (2001) sought to measure social presence through analyzing
online discussions. As touched on in Chapter 1, Rourke et al. identified three different
categories of social presence: affective responses, interactive responses, and cohesive
responses. They then developed twelve indicators that researchers could use to analyze
transcripts of CMC (primarily through content analysis). An example of these indictors
can be seen in Table 2.2 (see Appendix A for the complete list of indicators). Rourke et
al. developed these categories and indicators based on their previous work (Garrison,
Anderson, & Archer, 2000; Rourke, et al., 2001a), other literature in the field, and finally
their experience reading online transcripts.

51. 36
Table 2.2 Example of Social Presence Indicators
Category Indicators Definition of Indicators
Affective
Responses
Expression of
emotions
Conventional expressions of emotion, or
unconventional expressions of emotion,
includes repetitious punctuation,
conspicuous capitalization, emoticons
Use of Humor Teasing, cajoling, irony, understatements,
sarcasm
Self-Disclosure Presents details of life outside of class, or
expresses vulnerability
Rourke et al. (2001a) tested and measured the “efficacy and reliability” of their
categories and indicators by using them with participants in two graduate education
online courses. One single week from each course was identified, and all of the
discussion postings for those two weeks were analyzed. The first course had more than
twice the number of postings and words as the second course; as a result, in order to
compare the two, Rourke et al. (2001a) summed the raw number of instances and divided
by the total number of words and then multiplied it by 1000 to come up with a social
presence density score. They had high interrater reliability.
Rourke et al. (2001a), though, cautioned readers about generalizing their results
because their main purpose was to “develop and test the efficacy of a tool for analyzing
the social presence component of educational computer conferences” (Discussion
section) rather than to draw conclusions specifically about the samples in question. They
also acknowledged that they were still unclear whether all 12 indicators should be
weighted equally, as well as whether or not there was an optimal level of social presence.
In fact, Garrison mentioned in a round table presentation at the 2008 annual meeting of

52. the American Educational Research Association (AERA) that these indicators might need
37
to be revisited to ensure that they do not need to be revised (Arbaugh et al., 2008)
Tu and The Social Presence and Privacy Questionnaire
Tu (2002b) criticized early research on social presence (e.g., Short et al., 1976,
and even Gunawardena’s 1995 study) in which researchers adopted the same semantic
differential technique that simply had people respond to a bi-polar scale. Tu argued that
this technique is not adequate to measure one’s perception of social presence when it
comes to CMC. He also argued that the questionnaire used by Gunawardena and Zittle
(1997) failed to take into consideration different variables cited in the research (e.g.,
recipients, topics, privacy, task, social relationships, communication styles). As a result,
Tu (2002b) developed “The Social Presence and Privacy Questionnaire (SPPQ).”1 Tu
developed the SPQQ by using parts of Steinfield’s (1986) CMC attitude instrument and
Witmer’s (1997) work on privacy.
Tu used a panel of five qualified content experts to test the content validity of the
instrument. However, he did not elaborate on what made these content experts
“qualified.” He then used 310 inservice and preservice teachers to test the construct
validity. Five factors emerged from the factor analysis: social context, online
communication, interactivity, system privacy, and feelings of privacy; these five factors
accounted for 82.33% of the variance with Cronbach’s alpha values ranging from .74 to
.85. Tu acknowledged that online privacy had a weak correlation and therefore might not
need to be included as a dimension of social presence. However, he continued to use
1 In a different article, Tu (2002a) refers to the SPPQ as the CMC Questionnaire;
however, he tends to refer to it more often as the SPPQ and therefore SPPQ will be used
to refer to this instrument.

53. online privacy as a dimension of social presence in later studies (Tu & Corry, 2004; Tu &
38
McIsaac, 2002). Despite the strengths of his survey, Tu and McIsaac (2002) later
determined as the result of a mixed method study, using the SPPQ and a dramaturgy
participant observation qualitative approach, “there were more variables that contribute to
social presence” (p. 140) than previously thought. Therefore, Tu and McIsaac concluded
that social presence was more complicated than past research suggested. Appendix A
outlines the new variables identified by Tu and McIsaac. Specifically, they found that the
social context played a larger role than previously thought.
Among other things, the preceding literature illustrates what other researchers
have pointed out—that there is still little agreement on how to measure social presence
(Lin, 2004; Stein & Wanstreet, 2003). Just as Tu criticized how Gunawardena measured
social presence, others have criticized and modified Tu’s work (Henninger &
Viswanathan, 2004). Also, while social presence has been presented as a perceptual
construct, Hostetter and Busch (2006) point out that relying solely on questionnaires (i.e.,
self-report data) can cause problems because “respondents may be providing socially
desirable answers” (p. 9). Further, Kramer, Oh, and Fussell (2006) point out that self-report
data “are retroactive and insensitive to changes in presence over the course of an
interaction [or semester]” (p. 1). But at the same time, even the scale created by Rourke
et al. (2001a) has been modified by Swan (2003) and later by Hughes, Ventura, and
Dando (2007) for use in their own research.
During the past few years, researchers have focused less on studying social
presence by itself—opting instead to study social presence as one aspect of a CoI. As a
result and likely due to the difficulty of coding large samples, these researchers have

54. 39
focused almost predominantly on studying students attitudes toward the CoI as a whole
and each of the components of the CoI (i.e., social presence, teaching presence, and
cognitive presence). In 2008, a group of researchers came together to develop an
instrument to study the community of inquiry, called the Community of Inquiry
Questionnaire (see Arbaugh et al., 2008; Swan et al., 2008). Table 2.3 lists the part of the
Community of Inquiry Questionnaire used to assess students’ perceptions of social
presence in a CoI (see Appendix A for the entire instrument).
Table 2.3 Social Presence Dimension of the Community of Inquiry Questionnaire
Affective expression
14. Getting to know other course participants gave me a sense of belonging in the
course.
15. I was able to form distinct impressions of some course participants.
16. Online or web-based communication is an excellent medium for social
interaction.
Open communication
17. I felt comfortable conversing through the online medium.
18. I felt comfortable participating in the course discussions.
19. I felt comfortable interacting with other course participants.
Group cohesion
20. I felt comfortable disagreeing with other course participants while still
maintaining a sense of trust.
21. I felt that my point of view was acknowledged by other course participants.
22. Online discussions help me to develop a sense of collaboration
Over five years ago and before the work of Arbaugh et al., Russo and Benson
(2005) argued that researchers need “a multifaceted presence instrument, one that
examines presence more than single items and addresses the construct more by evaluating
specific behaviors rather than a global effect” (p. 60). And while Arbaugh et al. (2008)
hope that the Community of Inquiry Questionnaire is a step in that direction, as a whole

55. 40
their survey and the research in which it is used for the most part focuses on looking at
the CoI as a whole rather than at its parts (e.g., social presence).
In the end, though, the instrument that researchers use largely influences what
they find. Therefore, any study of social presence should at least acknowledge how its
methodology has been influenced by these early pioneers. Despite the varied
methodologies employed and some contradictions, some trends emerge when looking at
the research on social presence. The following section focuses first on the results of
some research on social presence and then on some recent research focused on social
presence in a CoI.
Research on Social Presence
Despite the differences previously noted, researchers have identified a number of
pedagogical implications—in most cases, benefits—of social presence. In the following
sections, the literature on social presence is summarized and synthesized around three
main themes: (a) social presence and student satisfaction, (b) social presence and
interaction, and (c) social presence and student learning.
Social Presence and Student Satisfaction
Over the years, a number of researchers have shown that there is a consistent
relationship between social presence and student satisfaction (Gunawardena, 1995;
Gunawardena & Zittle, 1997; Hostetter & Busch, 2006; Richardson & Swan, 2003; So &
Brush, 2008). While their conceptualization and methodology differ at times, most
researchers agree that social presence is a predictor of student satisfaction in CMC
environments, which in turn is a key component of online learning. More specifically, in
online learning environments student satisfaction has been connected to student

56. 41
persistence (Levy, 2007; Willging & Johnson, 2004). Levy (2007) has shown that
student satisfaction “is a major factor in students’ decision to complete or drop” online
courses (p. 198). Therefore, given the importance of student satisfaction, the following
section highlights a few of the main studies on social presence and student satisfaction.
In the 1980s and early 1990s, a number of researchers began investigating social
presence and computer-mediated communication (CMC) (e.g., Walther, 2002, 2004).
However, Gunawardena (1995; Gunawardena & Zittle, 1997) is perhaps the earliest, most
frequently cited, and foundational researcher of social presence and learning
environments using CMC. Gunawardena conducted two studies with Globaled
conference participants (Gunawardena, 1995; Gunawardena & Zittle, 1997). The studies
consisted of graduate students from different universities who attended the Spring 1992
and Fall 1993 Globaled computer conferences via a listserv.2 The participants in the
studies filled out questionnaires after they completed the conferences.
Gunawardena (1995) reported that, contrary to popular opinion, CMC could be
perceived as a social medium and that social presence could be cultivated. Further, she
stated that,
although CMC is described as a medium that is low in nonverbal cues and social
context cues, participants in conferences create social presence by projecting their
identities and building online communities. In order to encourage interaction and
collaborative learning, it is important that moderators of computer conferences
promote the creation of conducive learning environments. (p. 163)
Gunawardena and Zittle (1997), working from data collected from participants in the Fall
1993 conference, later reported that social presence was a strong predictor of student
2 It is important to highlight that the majority of the students in these studies completed
the online learning experience (i.e., the Globaled conference) as a component of a face-to-
face course; further, they took part in the conference via a listserv rather than a course
management system like Blackboard, WebCT, or eCollege.

57. satisfaction with computer conferences. They also found that students who felt a stronger
sense of social presence enhanced their socio-emotional expression (e.g., through the use
42
of emoticons) whereas those with a low sense of social presence did not. Gunawardena
and Zittle concluded that social presence (and as a result student satisfaction) depends on
what instructors and students do rather than simply the characteristic of a CMC medium.
Despite shortcomings of their research (e.g., small sample size, sample selection, course
format) as well as the fact that they caution readers not to generalize their results, the
work of Gunawardena (1995) and Gunawardena and Zittle (1997) is regularly cited—and
generalized—as foundational research on social presence and CMC.
Research conducted by Richardson and Swan (2003) is arguably less foundational
than the work of Gunawardena and Zittle but methodologically more sound. Richardson
and Swan (2003) conducted a study to investigate the relationship between students’
perception of social presence, perceived learning, and satisfaction with their instruction.
Their study consisted of 97 participants taking online courses at Empire State College, a
site purposefully chosen because of its nontraditional online program. Almost half of the
students in the sample stated that it was their first online course. Richardson and Swan
developed a survey based on Gunawardena and Zittle’s (1997) survey and used a
multiple regression to analyze the data collected from the survey.
Richardson and Swan (2003) found three things from their study. First, they found
that students with higher perceived social presence scores perceived they learned more
than students with lower scores, thus indicating that there is “a relationship between
students’ perceived social presence and students’ perceived learning” (p. 77). Second,
they found a link between student satisfaction with their instructor and perceived

58. 43
learning—which researchers have been finding in face-to-face settings for years. Third,
they found that students with high social presence scores “were highly satisfied with their
instructor” (p. 73). However, it is important to note that they did not find a relationship
between age or amount of college experience and social presence. Further, they
concluded that online learners found the social presence of faculty and students to be an
integral aspect of an online course.
Other researchers (Hostetter & Busch, 2006; Russo & Benson, 2005; So & Brush,
2008) have found a relationship between social presence and student satisfaction in
online learning environments as well. In fact, student satisfaction is the most consistent
finding across all studies of social presence and CMC.
However, like most findings on social presence, there always seems to be at least
one study that contradicts the findings of others. For instance, Joo, Lim, & Kim (2011)
recently sought to investigate the structural relationships between perceived level of
presence, perceived usefulness and ease of online tools, and learner satisfaction and
persistence at a South Korean online university (p. 1654). They administered two
different surveys resulting in 709 responses. While they found teaching presence had a
significant effect on both social presence and cognitive presence (which suggests how a
course is designed and facilitated effects social presence), they also found that contrary to
previous studies, social presence was not a significant predictor of satisfaction. Further
research though is needed to see if this study is an outlier or if perhaps students’
perceptions of social presence and its relationship to satisfaction is changing.
While student satisfaction does not equal student learning, it is a necessary
component of a successful learning environment. Further, online learning has a history of

59. 44
having a higher dropout rate than face-to-face courses (Levy, 2007) as well as being
characterized as involving more work than traditional face-to-face courses. Thus, it is
imperative for online instructors to recognize the important role student satisfaction can
play in online learning environments. If students are not satisfied, they will presumably
not log-on to their online course and therefore will not successfully complete their online
course or learn the required material. Another important component to student learning
online is interaction. The following section will address the relationship between
interaction and social presence.
Social Presence and Interaction
Interaction is a key component of any learning environment (Dunlap, Sobel, &
Sands, 2007). Interaction and online learning has specifically received a great deal of
attention over the years (Anderson, 2006; Anderson & Garrison, 1998; McIsaac, Blocher,
Mahes, & Vrasidas, 1999; Moore, 1989; Moore & Kearsely, 2005; Vrasidas & Glass,
2002; Wagner, 1994). Interaction has been defined in a number of ways. According to
Wagner (1994), interaction is simply “reciprocal events that require at least two objects
and two actions. Interactions occur when these objects and events mutually influence one
another” (p. 8). Interaction—that is, reciprocal events—can occur in many different
forms in an online environment.
Moore (1989) was the first to identify three main types of interaction in distance
education: (a) learner-to-content interaction, (b) learner-to-instructor interaction, (c) and
learner-to-learner interaction. Later researchers identified additional types of interaction
found in online learning environments: (a) teacher-to-teacher, (b) teacher-to-content, (c)
content-to-content, (d) learner-to-technology, and (e) teacher-to-technology (Anderson,

60. 45
2003, 2006; Anderson & Kuskis, 2007; Shank, 2004). Each of these is an important
component of any online learning environment. Furthermore, each of these types of
interaction can influence social presence. However, learner-to-instructor and learner-to-learner
interaction are the most germane; of the two, learner-to-learner interaction has
received the most attention.
Researchers have shown that learner-to-learner interaction is a critical component
in online learning (Richardson & Swan, 2003). Learner-to-learner interaction is
motivating and stimulating for students (Moore & Kearsley, 2005). Further, social
presence is directly related to learner-to-learner interaction (Tu, 2000). Students are
perceived as being there as a result of their online interactions with their peers; if they do
not interact with their peers and instructors, they are not perceived as being there or
connecting with their peers or instructors. Therefore, in this section, I will summarize a
few key studies about social presence and interaction.
Tu and McIsaac (2002) conducted a mixed methods study with 43 graduate
students in an online course. They found that social presence influences online
interaction. More specifically, they found that social presence is necessary for social
interaction. However, they also found that the quantity or frequency of participation
online did not directly relate to social presence. That is, interacting more did not
necessarily increase one’s social presence. Finally, they found that group size in
synchronous discussions influenced how much students interacted with others online.
As a result of their study, Tu and McIsaac (2002) argued that students need to
establish trust “before attaining a higher level of social presence” (p. 142). They also
found that informality helped increase social presence. But most importantly, they

61. 46
concluded that it is not the quantity but the quality of interactions online that make the
difference.
Like Tu and McIsaac, Swan and Shih (2005) also discovered some interesting
relationships between social presence and interaction. The participants in their study
came from four online graduate educational technology courses; 51 students completed
an online questionnaire [based on Richardson and Swan’s (2003) previous survey that
was based on the instrument developed by Gunawardena and Zittle]. After the survey was
completed, the 5 students with the highest and the 5 students with the lowest social
presence scores were identified, their postings were analyzed, and then they were
interviewed. Content analysis was used to explore the discussion postings using the
indicators developed by Rourke et al. (2001a).
Like Tu and McIsaac (2002), Swan and Shih found that students who interacted
more in the discussion forums were not necessarily perceived as having the most social
presence; rather, students who were more socially orientated, even if they interacted less
than others, were perceived as having greater social presence. Swan and Shih argued that
this supports the idea that perceived presence is not directly linked to how much one
participates online. Further, they found that students perceiving the most social presence
of others were also the ones who successfully projected their own presence into the
discussions. Swan and Shih concluded that students projected their presence online “by
sharing something of themselves with their classmates, by viewing their class as a
community, and by acknowledging and building on the responses of others” (p. 124),
rather than simply posting more than others. Therefore, this research suggests that the
quality of online postings matters more than the quantity when it comes to social

62. 47
presence. Finally, unlike earlier research (e.g., Richardson & Swan, 2003), Swan and
Shih found that the age of participants did matter. More specifically, they found that
younger students were more comfortable with online discussions than older students and
that students over the age of 45 did not bond well with other (younger) students (p. 123).
Even so, these differences could be due to the different samples; further, findings such as
these are likely to diminish over time, as more and more people of all ages spend time
online.
It is very possible that the occurrence of contradictory findings, like these about
age and social presence, arise because researchers have studied social presence in very
different contexts, with very different course formats, and very different groups of
students and instructors. Tu and McIsaac (2002) have already shown the important role
social context plays in social presence. Therefore, while age might diminish as a
significant variable over time, course format (e.g., hybrid vs. predominantly
asynchronous vs. predominantly synchronous), length of term, subject of study, and
students’ experience with online courses might continue to influence how social presence
is perceived, maintained, and enhanced in online learning environments.
Perhaps the most important research conducted on social presence has focused on
student learning. But just like previous research, researchers studying social presence
and student learning have found mixed and contradictory results.
Social Presence and Student Learning
Very few researchers have investigated the relationship of social presence and
student learning. That is, of the over 100 studies conducted on social presence, less than a
handful of studies have focused on student learning (Picciano, 2002; Wise et al., 2004)—

63. and the majority that have only focused on perceived learning (e.g., Richardson & Swan,
48
2003). This is not, though, simply a trend restricted to research on social presence and
online learning. Tallent-Runnels et al. (2006) have shown that generally speaking, “few
studies actually focus on instruction and learning online” (p. 116). This is most likely due
to the difficulty of measuring student learning (Fryshman, 2008). Because students are
complex creatures with vastly different backgrounds and experiences, it is often difficult
to determine what was learned specifically as the result of what happened in a given
course. As a result, most researchers who do focus on student learning actually
operationalize it as student performance on course assessments. This is true of Picciano
(2002) and Wise et al. (2004), the two researchers who have studied social presence and
student learning.
Picciano (2002) was one of the first to investigate social presence and student
learning. He was interested specifically in the relationship between social presence,
student interaction and performance, and student perceptions of social presence and
actual participation. Participants in Picciano’s study came from one completely online
asynchronous graduate education course. The 23 participants in the study were teachers
seeking a school administrator certification for the state of New York; they all had at
least five years teaching experience and a MA (Picciano, 2002). Only eight of the 23
students had taken an online course before. It is also important to note that the World
Trade Center was attacked during week two of this course; the attack on the World Trade
Center had a great impact on people across the world—but especially in New York.
Therefore, it is likely that the emotional distress felt by many Americans during this time

64. 49
could have impacted the results of this study and the social presence of the participants
online.
Picciano (2002) collected three different types of data. First, he calculated how
often students participated in the course discussions. However, interestingly, he chose
not to count “one line ‘me too’ postings and social messages” (p. 29); instead he wanted
to focus on “substantive comments.” Given the importance “me too” postings and
“social comments” can play in social presence, this decision likely skewed the results.
He also administered a student satisfaction survey—questions coming from both the
Inventory of Presence Questionnaire developed by the Presence Research Working
Group and Tu’s Social Presence and Privacy Questionnaire (SPPQ). Then, finally, he
collected scores from an exam and a written assignment. The exam was a multiple-choice
exam designed to explore 13 issues the course focused on. The written assignment was a
case study. Picciano created three groups—low, moderate, and high—of student
interaction, and then three groups of social presence (based on the survey). He then
calculated correlations of student interaction and the scores on the exam and the written
assignment as well as social presence and the performance on the exam and the written
assignment. Due to sample size, Picciano only used basic descriptive statistics (i.e.,
means and correlations) to analyze his data.
Regarding student perceptions of interaction and learning, Picciano (2002) found
that “there is a strong, positive relationship between student perceptions of their
interaction in the course and their perceptions of the quality and quantity of their
learning” (p. 28). Further, there was a positive relationship between student perceptions
of interaction, social presence, and their performance. However, he found that perceived

65. 50
performance and actual performance differed. To investigate actual student interaction
and performance, he found that actual student interaction had no relationship to the
performance on the exam but did have a relationship to the written assignment for the
highly interactive group. Finally, regarding social presence and performance, he found a
positive statistical relationship between social presence and the written assignment
(.5467) but not a positive statistically significant relationship between social presence and
the performance on the exam (-.3570).
Wise et al. (2004) also investigated student learning and social presence. Unlike
Picciano (2002), Wise et al. used an experimental research design that looked at how
social presence is related to learning in self-paced, one-to-one mentoring-supported,
online courses. That is, rather than using one online course that relies heavily on
student-to-student discussions, Wise et al. focused on self-paced online courses designed
for teacher professional development. These were self-paced courses typically completed
in a 12 week time period. However, participants in this study had to complete them in 6
weeks because it was an assignment for another course they were taking.
Twenty students taking a graduate course called Elementary and Secondary
School Curriculum took part in the study; half of the participants had teaching experience
and half did not. The students with teaching experience were randomly assigned to either
a high or low social presence condition; the students without teaching experience were
evenly distributed among the conditions (Wise et al.). For each condition, two instructors
were randomly assigned to five students (p. 255). The instructors had been trained on
social presence cues—which included the indicators developed by Rourke et al. (2001);

66. they varied the social presence cues they used based on the condition they were assigned
51
to.
The researchers physically went to students’ courses to solicit their participation.
Students then completed a pre-assessment survey which focused on demographics and
learning intentions. Then after the course was over, they had students complete a survey.
They did not include the survey in their article, and they never explicitly said how the
survey was constructed. They calculated message length as well as level of student social
presence by analyzing student messages with a type of content analysis. Then the final
assignment of the course—which was a lesson plan on integrating technology—was
assessed by two raters using a rubric.
They found that students in the high social presence condition replied with
messages twice as long as those of students in the low social presence condition. Further,
students in the high social presence condition tended to show a higher degree of social
presence in the content of their messages to the instructors. However, Wise et al. did not
find a statistically significant relationship between how students performed on the final
assignment of the course and the social presence condition. Further, unlike previous
studies, they did not find a significant effect between the condition and student
satisfaction or the condition and perceived learning. These findings led Wise et al.
(2004) to conclude the following:
Social presence impacts the atmosphere of the course as indexed by the
perceptions of the instructor and the nature of the interaction, but there is no
identifiable effect on the overall impact of the course as indexed by learning or
perceived learning, engagement, or satisfaction. (p. 262)

67. While Wise et al. try to make sense of their results, they never acknowledge how certain
issues—such as unique course format, limited time frame, lack of other students (to name
52
just a few)—might have skewed their results.
While Picciano (2002) and Wise et al. (2004) directly investigated social presence
and student learning (which they operationalize as student performance on specific
assignments), other researchers have investigated social presence and student perceived
learning. For instance, as mentioned earlier, Richardson and Swan (2003) conducted a
study in which they found that students who were identified with higher social presence
perceived they learned more than those with lower social presence. They also found a
relationship between student satisfaction with their instructor and perceived learning.
Russo and Benson (2005) also found a statistically significant relationship between
student perceptions of their own presence and the points they earned in a class. However,
Hostetter and Busch (2006) did not find a relationship between students’ perception of
presence and learning outcomes. Inconsistencies such as these, as well as those between
Picciano and Wise et al., suggest that the findings on social presence and student learning
are inconclusive.
Despite previous researchers’ efforts like those just described, researchers such as
Rourke and Kanuka (2009) have been critical of research on social presence and the CoI
as a whole for not spending enough time showing whether or not teaching presence and
social presence actually influence student learning. But despite some of the
inconsistences reported so far, research suggests social presence plays an important role
in online courses.

68. 53
Establishing and Maintaining Social Presence
Despite some mixed results, social presence appears to be related to student
satisfaction, student interaction online, students’ sense of community, as well as possibly
student learning. Therefore, as the number of students taking courses online increases
each year (Allen & Seaman, 2010), it is critical to better understand how social presence
is established and maintained in online learning environments.
The categories and indicators developed by Rourke et al. (2001a) and later
expanded upon by Swan (2003) can be seen as guidelines for establishing social
presence. For instance, if the expression of emotions, the use of humor, or self-disclosure
are indicators of social presence, then it is reasonable to expect that if one expresses his
or her emotions more, uses humor, and self-discloses information then he or she will be
able to establish social presence. Further, research suggests (e.g., Wise et al., 2004) that
social presence behaviors engender more social presence. Thus, indicators like those
listed in Table 2.4 should then be seen as guidelines for establishing social presence.
Table 2.4 Strategies to establish and develop social presence
Categories Indicators Strategies
Affective Emotions
Humor
Self-disclosure
Value
Paralanguage
Express emotions
Use humor
Self-disclose personal information
Express personal values
Use features of text, like emoticons to
express emotion

69. 54
Table 2.4 (con’t.)
Cohesive Greetings & salutations
Vocatives
Group reference
Social sharing
Self-reflection
Greet other students
Address students by name
Use inclusive pronouns like “we,” and “us”
Share personal information
Reflect on the course openly
Interactive Acknowledgement
Disagreement
Approval
Invitation
Personal advice
Refer directly to others’ postings
Agree or disagree with others’ postings
Express approval
Ask questions
Offer advice to peers
Note. From “Developing Social Presence in Online Course Discussions,” by K. Swan,
2003, in S. Naidu (Ed.), Learning and Teaching with Technology: Principles and
Practices (pp. 147-164). London: Kogan Page.
Synthesizing past literature (which included the work of Rourke et al.), Aragon
(2003) identified a similar list of strategies on how to establish social presence. Aragon
identified three components to establishing social presence online: (a) the role of course
design, (b) the role of the instructor, and (c) the role of the participants (see Table 2.5).
Aragon’s three components to establishing social presence are similar to the three types
of presence—social presence, teaching presence, and cognitive presence—of the CoI
framework developed by Garrison et al. (2000). Aragon’s strategies for instructors
though are not simply what Garrison et al. would call “teaching presence” strategies in
the CoI framework.

70. 55
Table 2.5 Strategies to Establish and Maintain Social Presence
Course Design Instructors Participants
Develop welcome
messages
Include student profiles
Incorporate audio
Limit class size
Structure collaborative
learning activities
Contribute to discussion
boards
Promptly answer email
Provide frequent feedback
Strike up a conversation
Share personal stories &
experiences
Use humor
Use emoticons
Address students by name
Allow students options for
addressing the instructor
Contribute to discussion
boards
Promptly answer email
Strike up a conversation
Share personal stories and
experiences
Use humor
Use emoticons
Use appropriate titles
Note. From “Creating Social Presence in Online Environments,” by S. R. Aragon,
2003, in, New Directions for Adult and Continuing Education, 100, (pp. 57-68).
Certain strategies such as “sharing personal experiences and stories” are more of a
type of social presence strategy used by instructors. This is important because the CoI
traditionally does not differentiate between how a student and how an instructor
establishes his or her social presence. Swan and Shih (2005) though pointed out that there
are some differences between the two. And Nippard and Murphy (2007), who sought to
investigate how teachers and students manifest social presence in synchronous web-based
secondary classrooms, found that in their study students and instructors did in fact engage
in different social presence behaviors.

71. Some question whether there may be an optimal level of social presence (Garrison
& Anderson, 2003; Wise et al., 2004). For instance, Garrison and Anderson posit that too
little social presence can prevent a learning community from forming but too much social
56
presence may reduce student learning by discouraging critical discourse (Garrison &
Anderson, 2003). Conjectures such as these might suggest that while it might be
important to focus on establishing social presence early in a course, it might be less
important later on in the course. Dunlap and Lowenthal (2010), though, have argued
based on their experience that it is important to continue to maintain social presence
throughout the duration of an online course.
Mixed results arise, though, on how social presence is maintained throughout the
duration of a course. For instance, the research of Rourke et al. (2001a) suggests that
purely social communication decreases over time. However, Stacey (2002) later found in
her study that purely social communication did not decrease but actually increased over
time; as a result, she concluded “that social relationships require continuation of social
presence factors through a much longer period, as one semester is only the beginning of
group formation online” (p. 150). But then to confuse matters even more, Swan (2003)
later found that “although the use of affective indicators mirrored the general flow of the
course discussions as the course progressed, cohesive indicators declined in importance,
while the importance of interactive indicators increased” (pp. 161-162). This possibly
suggests that different types of social presence behaviors serve different functions and
vary across time. For instance, Lomicka and Lord (2007)—who studied how foreign
language graduate students established and maintained social presence through weekly

72. 57
journal activities—found that the type of task influenced the type of presence used by
students.
Some Gaps in the Literature
Despite the popularity of social presence research, a number of gaps in the
literature remain. As researchers and practitioners continue to rely on theories and
research on social presence to influence their practice, it is imperative to address some of
these gaps. In the following section, I will briefly describe a few of these gaps that future
research needs to address.
First, as indicated throughout this chapter, previous researchers have found mixed
and contradictory results. For instance, some studies have found a strong relationship
between student satisfaction and social presence (Gunawardena, 1995; Gunawardena &
Zittle, 1997; Richardson & Swan, 2003), but other studies have not (Joo, Lim, & Kim,
2011; Wise et al., 2004). Some studies have found a relationship between social presence
and student performance (whether perceived or actual) (Picciano, 2002; Richardson &
Swan, 2003; Russon & Benson, 2004) while others have not (Hostetter & Busch, 2006;
Wise et al., 2004). Finally, some studies have found that social presence changes over
time (Rourke et al., 2001a; Swan, 2003) while others have not (Lomicka & Lord, 2007;
Stacey, 2002). Recently, Akyol, Vaughan, and Garrison (2011) sought to investigate
how time effects the development of a CoI. They studied the CoI in a 6 week and a 13
week online course. They reported that an “independent samples t-test revealed
statistically significant differences between the short and long-term courses on affective
communication . . . and group cohesion” (Akyol et al., p. 235). More specifically and to
their surprise—because conventional wisdom suggests that more time is needed to

73. 58
develop group cohesion—they found that students’ messages in the 6 week course
included more group cohesion indicators than the students’ messages in the 13 week
course. At the same time, there were more affective indicators in the 13 week course than
the 6 week course.
Second, conclusions are drawn and assumptions made about the nature of social
presence from research conducted on vastly different types of online courses. For
instance, researchers have failed to recognize how the socio-cultural context and course
format effects social presence. For instance, researchers (e.g., Tu, 2001; Wise et al.,
2004) have conducted a number of studies on social presence in which the teacher had
face-to-face meetings with students; meetings like these will have likely impacted the
development and perception of social presence. This is not to suggest that meeting face-to-
face with online students is a poor decision, but rather that it can influence students’
perceptions and therefore should only be compared to other instances where faculty meet
face-to-face with their students.
In other cases, researchers have studied social presence in non-traditional courses.
For instance, Gunawardena (1995; Gunawardena & Zittle, 1997) studied social presence
in a computer conference administered through a listserv, Tu (2001; 2002a) studied social
presence in face-to-face courses using CMC as well as televised courses, and Wise et al.
(2004) studied social presence in six week self-paced courses. While it is important to
understand how social presence is developed and maintained in these nontraditional types
of online courses, it is even more important to recognize how social presence might
change across a variety of different online learning formats and contexts. For instance,
Lowenthal, Wilson, and Parrish (2009) have argued about the importance context plays in

74. 59
online learning. And both Arbaugh, Bangert, and Cleveland-Innes (2010) and Gorsky,
Caspi, Antonovsky, Blau, and Mansur (2010) have recently found some interesting
differences in students’ perceptions of the CoI across academic disciplines.
Third, the only relationship found between social presence and actual student
learning (as opposed to perceived student learning) comes from a study that was
conducted in New York City during the attack on the World Trade Center (Picciano,
2002); there is reason to believe that a tragic event such as this could have skewed the
results. Further, Picciano (2002) only found a statistical relationship with one of his
measures of student learning and social presence. However, on the other hand—as
already mentioned—a number of researchers (Picciano, 2002; Richardson & Swan, 2003;
Russon & Benson, 2004) have found a relationship between social presence and
perceived learning. Differences and issues like this illustrate that questions remain about
the relationship of student learning—both actual and perceived—and social presence.
Fourth, and finally, the majority of past research on social presence has heavily
relied on survey data to study social presence. Studying students’ perceptions of social
presence is important. However, relying only on self-report data can be problematic
because students might be simply providing socially desirable answers (Hostetter &
Busch, 2006, p. 9). Further, self-report data tend to be collected at one period during a
semester and therefore cannot show change over time (Kramer, Oh, & Fussell, 2006).
Unfortunately, though, compared to studies using only self-report data, few researchers
(e.g., Delfino & Manca, 2007; Hughes et al., 2007; Rourke et al., 2001a; Swan, 2003)
have actually analyzed online discussions when studying social presence. But if
researchers want to understand better how social presence develops, is maintained, and

75. 60
changes over a course, they must begin to look at what is “said” and done in threaded
discussions—the primary avenue for interaction.
Chapter Summary
Researchers and practitioners alike seem fascinated by the concept of social
presence. I have found over a 100 articles on the subject. However, like most research on
online learning (Bernard et al, 2004; Tallent-Runnels, 2006), research on social presence
and online learning is of mixed quality. Even though initial research suggests that social
presence is related to student satisfaction, student interaction, and student learning, many
questions remain. In the next chapter, Chapter 3, I will outline the methods used for this
study. Then in Chapter 4, I go over the results of the study and I conclude in Chapter 5
with a discussion of the results.

76. 61
CHAPTER 3
METHOD
The purpose of this study is to explore the phenomena of social presence in an
online graduate level education course at the University of Colorado Denver. To
accomplish this, I utilized a mixed research methods approach that employed both
quantitative and qualitative methods to understand better social presence. In the following
chapter, I elaborate on the methods used for this study.
Research Questions
Research questions help narrow the focus of a study by providing a framework,
setting boundaries, and giving rise to the type of data that will be collected (Cresswell &
Plano Clark, 2007). The following research question guided this exploratory study: How
does social presence manifest in an asynchronous, online graduate-education course?
Research Design
Mixed methods research has become popular over the past few years (Leech &
Onwuegbuzie, 2006). Around the same time, researchers of CMC (e.g., Goldman,
Crosby, Swan, & Shea, 2005; Gunawardena, Lowe, & Anderson, 1997; Hiltz & Arbaugh,
2003) began arguing about the importance of using multiple methods when studying the
complexity of asynchronous learning environments. However, to date the majority of
research on social presence has utilized either a quantitative or qualitative approach,
which conceivably limits researchers interpretation of the data and understanding of the
phenomenon. I utilized a mixed methods research design in this study. The purpose of
using a mixed methods approach was to facilitate the richness of data and expand the

77. 62
interpretation of the findings (Collins, Onwuegbuzie, & Sutton, 2006; Onwuegbuzie &
Leech, 2004), as well as to answer the research question that guided this study.
Sample
Researchers differentiate between sampling schemes and sampling designs
(Onwuegbuzie & Collins, 2007; Onwuegbuzie & Leech, 2007b). Thus, I elaborate below
on the sampling scheme and the sampling design used in this study.
Sampling Scheme
The goal of this study was to gain insights into a phenomenon (i.e., social
presence) rather than to generalize findings to a population. In situations like these,
Onwuegbuzie and Collins (2007) argue that a purposeful sample should be used.
Therefore, a non-random (non-probability) criterion sampling scheme was used in this
study.
A section of EDLI 7210 Educational Policy Making for a Democratic Society—
which was taught in the spring of 2007 at the University of Colorado Denver—was
identified as an appropriate sample for this study. This course was selected for a number
of reasons. First, this course was a fully online course. While it is helpful to study social
presence in hybrid or televised courses, the focus of this study is on the nature of social
presence in fully online courses. Second, this course was an asynchronous, instructor-facilitated,
online course—the most popular type of online course in higher education.
And third, this was an education course. Recognizing that CMC is always socially
situated (Herring, 2004), the goal of this study was to study social presence in an
education setting—like the majority of previous research on social presence (Lowenthal,
Lowenthal, & White, 2009).

78. 63
Nineteen graduate students were enrolled in the course. The following course
description describes the basic focus of the course:
This course examines the role and impact of policy and policy processes in
educational organizations. Models will be developed to analyze the nature of
policy, how policy processes work, conceptualizations of and research on these,
and their implications for improving educational practices to benefit student
learning and other organizational behaviors and outcomes. The study of policy
and policy processes will be facilitated by several activities that will familiarize
you with various perspectives on, models of, and research about the initiation of
policy issues, the processes of implementation and evaluation, and their outcomes
and effects. Collaboration, group work, and research are emphasized.
The learning objectives of the course are:
1. To read critically a variety of works on policy-making processes and outcomes
2. To develop skills as a policy analyst and advocate
3. To develop appreciation for and use of various policy models, policy research,
and policy effects
The following are the main assignments of the course (see Table 3.1 for a description of
each assignment):
• Reading logs
• A policy critique
• An observation
• A book review
• A personal-professional task
• A small-group project
• Online interactivity and quality of work.

79. 64
Table 3.1 Assignment Descriptions
Individual and Group Assignments
Individual Assignments
Policy Critique (8.47% 11.8% of course grade): Write a five page (or less) double-spaced
paper assessing the goals (intended), trends, conditions, projections,
alternatives relative to a local, regional, or state educational policy.
Observation (8.47% 11.8% of course grade): First observe and analyze a policy
process (school board meeting, city: council, state legislature, county commission).
Then write a 2-3 page analysis of their observation.
Book Review (8.47% 11.8% of course grade): Select a book related to policy studies
and write a 3-5 page review.
Online Interactivity and Quality of Work (16.95% of course grade): Login regularly,
take part in threaded discussions, and produce quality work.
Group Assignments
Reading Logs (15.25% of course grade): Discuss readings in small groups Then write
nine reading logs summarizing the readings and posing questions for the instructor.
Personal-Professional Task (12.7% of course grade): In a group of two, take some
risk and discuss deeply important individual personal-professional goals. Analyze
trends that facilitate or impede each other’s goal achievement. Then write a 3-5 page
paper summarizing his or her partner’s self analysis and plan.
Small-Group Project (23.73% of course grade): Working in small groups of 4-5
students, study a policy at the local, state, or national level. Then write a 10-15 page
double-spaced group paper employing a qualitative approach to collecting data that
informs their critical analysis of the policy.
In addition to the assignments, a number of different types of discussions were
conducted in the course. The most active discussions were the “General Discussion
Forum,” “Reading Groups,” “Reading Log Discussion Forum,” “Pairs,” and finally
“Project Groups” (which everyone but the “General Discussion Forum” was directly tied
to key assignments).

80. 65
Sampling Design
Social presence researchers who study online course discussions historically only
analyze a small sample of course discussions. For instance, Rourke et al. (2001a) only
analyzed one week of discussions in two different courses in their foundational study,
which consisted of 134 messages or 30,392 words. Swan (2003) analyzed 235 posts with
an average number of words per posting at 82.4 (which she explained was 10% of the
entire courses discussions). And then Hughes et al. (2007) analyzed three different groups
of students with a total of 974 messages or 63,655 words.
For this study, I chose to analyze every threaded discussion in the course using
content analysis, which consisted of 1,822 posts or 160,091 words (see Table 3.2). Then
based on the results of the content analysis, two specific threaded discussions (which
span multiple weeks of the class) were identified—one with the highest social presence
indicators (which was Pair 9) and one with the lowest social presence indicators (which
was Reading Group E)—and analyzed using constant comparison analysis in an effort to
explore better the phenomenon of social presence. It is important to note that for the
purpose of this study, all discussion forums in the course are referred to as “threaded
discussions”—regardless of their purpose, the level of dialogue, or the amount of
interaction between the instructor and the students.

81. 66
Table 3.2 Threaded Discussions Raw Data*
Discussion Name # of Participants # of Posts # of Words
Virtual Office 7 44 2560
General—Syllabus 14 48 3294
General—Groups 6 14 639
General—Independent Work 3 3 155
General—Individual Work 2 2 84
Adult Learning Discussion Forum—
7 12 456
Your Learning
Adult Learning Discussion Forum—
Questionnaire #1
3 3 221
A: Reading Group A 4 125 7828
B: Reading Group B 5 132 11677
C: Reading Group C 4 95 8452
D: Reading Group D 4 109 12562
E: Reading Group E 5 40 5235
F: Reading Group F 4 106 10916
G: Reading Group G 5 103 8116
Pair 1 3 32 2028
Pair 2 3 40 6222
Pair 3 3 45 3000
Pair 4+ 4 6 248
Pair 5 3 30 2232
Pair 6 3 28 1453
Pair 7 3 26 2687
Pair 8 3 21 3658
Pair 9 3 15 2909
Pair 10 2 22 2129
Plus Delta Week2 8 13 866
Plus Delta Week 3 8 22 2375
Plus Delta Week 4 2 2 299
Plus Delta Week 5 2 2 109
Plus Delta Week 6 3 3 234
Project Group 1 5 109 12673
Project Group 2 5 180 15322
Project Group 3 5 138 8404
Project Group 4 5 113 6791
Project Group 5 4 126 12380
Reading Log 1 5 12 1364
Reading Log 3 1 1 513
Total 156 1822 160,091
*Note. The discussion names were copied exactly as they were worded in the online
course. If a discussion did not have any posts (e.g., Reading Log 2), it was not listed.

82. 67
Data Collection
The data for this study were collected from the asynchronous threaded discussions
from an online course taught in eCollege—a learning management system used at the
University of Colorado Denver. While the course was taught in 2007, an archived copy
of the course is stored in eCollege. Archived copies of course discussions, like these,
“provide readily accessible records of the evolution of social relationships in online
classes” (Goldman, Crosby, Swan, & Shea, 2005, p. 109). The course discussions were
copied from eCollege to Microsoft Word; each discussion was saved as its own file.
Student names were replaced with pseudonyms, and the files were imported into NVivo
8.
Data Analysis
Initially when researchers began studying online discussions, they focused on the
frequency of participation (Henri, 1992). In fact, researchers have only relatively recently
begun to move beyond the basics (e.g., frequency of student participation, the level of
interaction, and message length) to focus instead on studying the content of messages
online (Pena-Shaff & Nicholls, 2004). When researchers began focusing on analyzing
the content of messages, they turned to content analysis (De Wever, Schellens, Valcke, &
Keer, 2006). In this study, though, I used three types of analysis to analyze the data: (a)
word count, (b) content analysis, and (c) constant comparison analysis (Leech &
Onwuegbuzie, 2007; see Table 3.3).

83. 68
Table 3.3 Overview of Data Analysis
Research Question: How does social presence manifest in a graduate education
asynchronous online course?
Data Analysis Type of Data Purpose of Results
• Word Count
(Quantitative)
• Content Analysis
(Quantitative)
• Constant
Comparative
Analysis
(Qualitative)
• All course
discussions
• All course
discussions
• One discussion thread
with high social
presence & one with
low social presence
• Explore the frequency of
top words used
• Explore the presence and
frequency of categories
and indicators of social
presence.
• Identify codes, groups,
and themes in the data
missed by content
analysis.
Word Count
Traditionally, word count involves identifying deductively a word or words from
the literature on a subject or inductively identifying from the data specific words that
seem out of place or hold special meaning and then counting the frequency of these
words. For the purpose of this study, word count was used solely as a way to initially
explore the data primarily by looking for the frequency of and more importantly the type
of words used in the online discussions. NVivo 8 can quickly and efficiently calculate
word counts.
Word count is an effective initial way to analyze data by exploring the occurrence
of words in a data set. The assumption with word counts, according to Leech and
Onwuegbuzie (2007), “is that more important and significant words for the person will be
used more often” (p. 568). However, it is important to acknowledge that word count has
some limitations (e.g., it can decontextualize a word and its meaning) (Leech &

84. Onwuegbuzie). Therefore, word count should not be used as the only method to analyze
data. And for this study, it was used solely as an initial method to identify if certain types
69
of words (e.g., student names or greetings and salutations which are indicators of social
presence) were used more than others.
Content Analysis
In the social sciences, content analysis has been the leading method used by
researchers to analyze text (Carley, 1993). Content analysis is understood and defined in
a number of different ways. For instance, Berelson (1952) defined it as “a research
technique for the objective, systematic, quantitative description of the manifest content of
communication” (p. 519). But then Holsti (1969) defined it as “any technique for making
inferences by objectively and systematically identifying specified characteristics of
messages” (p. 14). Finally, Carley (1993) explains that “content analysis focuses on the
frequency with which words or concepts occur in texts or across texts” (p. 81).
Regardless of how one defines it, the purpose of content analysis “is to reveal
information that is not situated at the surface of the transcripts” (De Wever, Schellens,
Valcke, & Van Keer, 2006, p. 7).
For this study, I followed the following five steps identified by Herring (2004):
1. The researcher formulates a research question and/or hypotheses
2. The researcher selects a sample
3. Categories are defined for coding
4. Coders are trained, code the content, and the reliability of their coding is
checked
5. The data collected during the coding process are analyzed and interpreted.

85. 70
I turned to the literature at step number three to identify categories for coding. I first
looked at the broad categories of the CoI framework developed by Garrison et al. (2000).
Garrison et al. identified three categories of social presence—namely, emotional
expression, open communication, and group cohesion. At that time, they only identified
some possible examples of indicators for each category (see Table 3.4). I then referred to
the work of Rourke et al. (2001a)3 in which Garrison and his colleagues more fully
developed the categories and indicators of social presence.
Table 3.4 Original Social Presence Categories and Example Indicators
Element Category Examples of Indicators
Social Presence Emotional Expression
Open Communication
Group Cohesion
Emotions
Risk-free expression
Encouraging collaboration
Rourke et al. changed the names of the categories from Emotional Expression to
Affective Responses, Open Communication to Interactive Responses, and Group
Cohesion to Cohesive Responses. They also identified specific indicators for each
category of social presence as well as definitions of each indicator (see Table 3.5).
3 Please note that some uncertainty exists regarding the original date of Rourke et al.’s
article entitled “Assessing Social Presence in Asynchronous Text-based Computer
Conferencing.” I personally have a hard copy in which the publication is listed as 1999
and another as 2001. Researchers tend to cite it both ways. I contacted Liam Rourke to
get some clarification but he simply replied that he was not sure but that he thought 2001
might be correct. I reference it as 2001 for the purpose of this study.

86. 71
Table 3.5 Rourke et al.’s Categories and Indicators of Social Presence
Category Indicators Definition of Indicators
Affective
Expression of
Responses
emotions
(originally
“Emotional
Expression”)
Conventional expressions of emotion, or
unconventional expressions of emotion,
includes repetitious punctuation,
conspicuous capitalization, emoticons
Use of Humor Teasing, cajoling, irony,
understatements, sarcasm
Self-Disclosure Presents details of life outside of class, or
expresses vulnerability
Interactive
Responses
(originally
“Open
Communication”)
Continuing a Thread Using reply feature of software, rather
than starting a new thread
Quoting from Other
Messages
Using software features to quote others
entire message or cutting and pasting
sections of others’ messages
Referring explicitly
to other messages
Direct references to contents of others’
posts
Asking questions Students ask questions of other students
or the moderator
Complimenting,
expressing
appreciation
Complimenting others or contents of
others’ messages
Expressing
agreement
Expressing agreement with others or
content of others’ messages
Cohesive
Responses
(originally
“Group
Cohesion”)
Vocatives Addressing or referring to participants by
name
Addresses or refers
to the group using
inclusive pronouns
Addresses the group as we, us, our,
group
Phatics/Salutations Communication that serves a purely
social function; greetings, closures
Note. From “Assessing Social Presence in Asynchronous Text-based Computer
Conferencing,” by L. Rourke, D. R. Garrison, and W. Archer, 2001a, in Journal of
Distance Education, 14.
Swan (2003), however, later made some changes to the list of indicators—namely
Swan simplified the interactive indicators but elaborated on the affective indicators (see

87. 72
Table 3.6). Finally, Hughes et al. (2007) replicated Rourke et al.’s (2001a) work but
apparently were unaware of Swan’s previous study. They too made some changes to the
indicators of social presence originally developed by Rourke et al. but some that were
different than Swan. Faced with this evolution of social presence indicators (see Table
3.6), I decided to integrate the changes both Swan and Hughes et al. made to the social
presence indicators (see Table 3.7). I used this initial combined list of indicators during
the first training session with one of two coders. During the training sessions, it became
apparent that the list of indicators needed to be amended.
Table 3.6 Evolution of the Indicators of Social Presence
Rourke et al. (2001a) Swan (2003) Hughes et al. (2007)
Categories & Indicators
Categories &
Indicators
Categories & Indicators
Affective Responses Affective Responses Affective
Expression of emotions
Use of Humor
Self-Disclosure
Paralanguage
Emotion
Value
Humor
Self-Disclosure
Expression of emotion
Use of Humor
Self-Disclosure
Interactive Responses Interactive Responses Interactive
Continuing a Thread
Quoting from Other
Messages
Referring explicitly to
other messages
Asking questions
Complimenting,
expressing appreciation
Expressing agreement
Acknowledgement
Disagreement
Approval
Invitation
Personal Advice
Referring to other’s
messages
Asking Questions
Complimenting,
expressing appreciation
Expressing Agreement
Cohesive Responses Cohesive Responses Cohesive
Vocatives
Addresses or refers to the
group using inclusive
pronouns
Phatics / Salutations
Greetings &
Salutations
Vocatives
Group Reference
Social Sharing
Self-reflection
Vocatives
Expresses group
inclusivity
Phatics / Salutations
Embracing the Group

88. 73
Table 3.7 Swan & Hughes et al. Combined List of Categories and Indicators of
Social Presence
Category & Indicator Definition (Swan) Criteria (Hughes)
Affective Responses
Paralanguage Features of text outside formal
syntax used to convey emotion
(i.e., emoticons, exaggerated
punctuation or spelling)
Emotion Use of descriptive words that
indicate feelings (i.e., love, sad,
hate, silly); conventional or
unconventional expression of
emotions
Refers directly to an emotion
or an emoticon. Use of
capitalization only if
obviously intended
Value Expressing personal values,
beliefs, and attitudes
Humor Use of humor—joking, teasing,
cajoling, irony, sarcasm,
understatement
Only code if a clear
indication that this is meant
to be funny, e.g., extra
punctuation or an emoticon
Self-Disclosure Sharing personal information,
expressing vulnerability or
feelings
An expression that may
indicate an emotional state
but does not directly refer to
it. Uncertainty, non
comprehension
Interactive Responses
Acknowledgement Referring directly to the
contents of others’ messages;
quoting from others’ messages
agreement
Explicit or implicit
recognition that another
message has been the
motivation for this message
Agreement /
Disagreement
Expressing agreement or
disagreement with other’s
messages
Expressing agreement with
each other or contents of
messages
Approval Expressing approval, offering
praise, encouragement
Invitation Asking questions or otherwise
inviting response. Students ask
questions of each other or
moderator
Personal Advice Offering specific advice to
classmates
Complimenting,
expressing
appreciation
Complimenting, expressing
appreciation
Complimenting or showing
appreciation of each other or
contents of messages

89. 74
Table 3.7 (con’t.)
Cohesive Responses
Greetings &
Salutations /
Phatics
Greetings, closures.
Communication that serves a
purely social function
Vocatives Addressing or referring to
classmates by name
Group Reference /
inclusivity
Referring to the group as “we,”
“us,” “our.” Addresses the
group as a possessed or as a
whole
Any reference to the
group with a possessive
pronoun
Social Sharing Sharing information unrelated
to the course
Not really about yourself
but more of a social
response
Self-reflection Reflection on the course itself,
a kind of self-awareness of the
group
Embracing the
Group
Revealing life outside the
group
Any expression that lets
the group know about the
circumstance of the
author but does not make
author vulnerable
For instance, under the affective category, the indicator of value was eliminated
because it was nearly impossible for two coders to reliably identify value. Further, given
the content of the course, nearly every other discussion posting appeared to have a “I
think” or “I feel. . . ” statement (which were the examples originally provided by Swan).
Under the interactive responses, the indicators of approval and personal advice were
eliminated. Approval was eliminated because of its overlap with complimenting/
expression appreciation and personal advice was difficult to identify thus complicating
reliability. Finally, under the category of cohesive responses, social sharing and self-reflection
were eliminated. Social sharing and embracing the group overlapped and self-

90. reflection was difficult to identify. The following coding sheet in Table 3.8 was used for
75
the content analysis.
Table 3.8 Coding Sheet Used for Content Analysis
Category &
Indicator
Definition
(Swan)
Criteria Examples
Affective Responses
Paralanguage
(PL)
Features of text
outside formal
syntax used to
convey emotion
(i.e., emoticons,
exaggerated
punctuation or
spelling)
Someday……; How
awful for you ;
Mathcad is definitely
NOT stand along
software;
Absolutely!!!!!
Emotion
(EM)
Use of descriptive
words that
indicate feelings
(i.e., love, sad,
hate, silly);
conventional or
unconventional
expression of
emotions
Refers directly to
an emotion or an
emoticon. Use of
capitalization only
if obviously
intended
When I make a spelling
mistake, I look and feel
stupid; I get chills when
I think of . . . I am
scared; This is fun;
Sorry this is such a lame
email; Hope you are
OK; I am pleased that
Humor (H) Use of humor—
joking, teasing,
cajoling, irony,
sarcasm,
understatement
Only code if a
clear indication
that this is meant
to be funny, e.g.,
extra punctuation
or an emoticon
God forbid leaving your
house to go to the
library
I’m useless at computers
but will this make me a
bad nurse??? Ha Ha ;
LOL
Self-
Disclosure
(SD)
Sharing personal
information,
expressing
vulnerability or
feelings
An expression that
may indicate an
emotional state but
does not directly
refer to it;
Uncertainty, non-comprehension
I sound like an old lady;
I am a closet writer; We
had a similar problem.
I’m not quite sure how
to . . .; This is strange; I
don’t understand how; I
don’t’ know what that
means; As usual I am
uncertain; It’s all too
much . . .; Website???
Help!!!!

91. 76
Table 3.8 (con’t.)
Interactive Responses
Acknowledgement
(AK)
Referring directly
to the contents of
others’ messages;
quoting from
others’ messages
agreement;
Reference to
others’ posts
Explicit or
implicit
recognition
that another
message has
been the
motivation for
this message
Those ‘old machines’
sure were something;
we won by a landslide
– ‘landslide’ (next
response)So what
you’re saying is . . .; I
thought that too . . . For
me the question meant .
. .;
Agreement /
Disagreement
(AG)
Expressing
agreement or
disagreement
with others’
messages
Expressing
agreement with
each other or
contents of
messages
I’m with you on that; I
agree; I think what you
are saying is right. I
think that would be a
good plan; I think your
suggestion is good
Invitation (I) Asking questions
or otherwise
inviting response.
Students ask
questions of each
other or
moderator
Any suggestions?;
Would you describe
that for me, I am
unfamiliar with the
term. Does anybody
know . . .?
Expressing
Appreciation (EA)
Showing
appreciation of
each other
Showing
appreciation or
approval of
each other or
contents of
messages or
complimenting
You make a good
point; Right on; Good
luck as you continue to
learn
I like your briefing
paper . . .; It was really
good;
Cohesive Responses
Greetings &
Salutations /
Phatics (GS)
Greetings,
closures.
Communication
that serves a
purely social
function
Hi Mary; That’s it for
now, Tom Hi; Hey;
Bye for now;
Vocatives (V) Addressing or
referring to
classmates by
name
You know, Tamara, . .
.; I totally agree with
you Katherine Sally
said that . . .

92. 77
Table 3.8 (con’t.)
Cohesive Responses (con’t.)
Group
Reference /
inclusivity
(GR)
Referring to the
group as ‘we’,
‘us’, ‘our’.
Addresses the
group as a
possessed or as a
whole
Any reference to
the group with a
possessive pronoun
We need to be
educated; Our use of
the Internet may not be
free. We need some
ground rules;
The task asks us to . . .
Embracing
the Group
(EG)
Revealing life
outside the group
that is not
emotional or
expressing
vulnerability or
feelings. Also that
isn’t related to the
course
Any expression that
lets the group know
about the
circumstance of the
author
The kids are asleep
now; I’m a
physiotherapist;
It’s raining again; It’s
4am—I’m off to bed;
Constant Comparison Analysis
Constant comparison analysis was the final type of analysis conducted on the
threaded discussions. Constant comparison analysis—a specific type of comparative
analysis—is a general method used in social science research that traces back to the work
of Glaser and Strauss and their development of grounded theory (1967). While
researchers like Krathwohl (2004) and Creswell (1998, 2008) approach constant
comparison analysis from only a grounded theory perspective, it is not restricted to a
grounded theory or inductive approach (Leech & Onwuegbuzie, 2007). Leech and
Onwuegbuzie (2007) explain that constant comparison analysis can be conducted
inductively, deductively, or abductively.
Constant comparison analysis is useful when trying to explore and understand the
big picture of a phenomenon (e.g., social presence). In fact, it is one of the most
commonly used types of qualitative analysis (Leech & Onwuegbuzie, 2007). However,

93. researchers of CMC rarely use it to analyze online course discussions, most likely due to
78
the time involved to conduct this type of analysis. It was used in this study to dig more
deeply into the threaded discussions to understand better the nature of social presence. To
conduct constant comparison analysis, I read the entire threaded discussion and
partitioned each meaningful unit into small chunks. I then labeled each chunk with a code
while constantly comparing new codes with previous ones. I then grouped the codes
together. Once I grouped the codes together, I identified themes that emerged from the
data. Figure 3.1 outlines the steps I took with examples from a previous study I
conducted.
Step 1. Read the discussion post
Hello everyone! I love the educational environments you have created this week. Educators and
students should always be the ones who create our schools. It is inspirational to see so many of you
create from the schools you have been in or are currently in. Thanks for your creativity! Dr. Deb C.
Step 2. Chunk the discussion post into meaningful units
[Hello everyone!] [I love the educational environments you have created this week.] [Educators
and students should always be the ones who create our schools.] [It is inspirational to see so many
of you create from the schools you have been in or are currently in.] [Thanks for your creativity!]
[Dr. Deb C]
Step 3. Code each meaningful unit while constantly comparing new
codes with previous codes
[Hello everyone!] GREETING [I love the educational environments you have created this week.]
POSITIVE FEEDBACK [Educators and students should always be the ones who create our
schools.] ELABORATION / CLARIFICATION [It is inspirational to see so many of you create
from the schools you have been in or are currently in.] POSITIVE FEEDBACK [Thanks for your
creativity!] POSITIVE FEEDBACK [Dr. Deb C] CLOSING REMARK

94. 79
Step 4. Make a list of the codes and group the codes
Codes Grouping of codes
Closing remark
Course logistics
Directions
Directions
Positive feedback
Writing style
Greeting
Number of students
Questioning
Teacher request
Answering question
Colorado law
Elaboration / clarification
Writing style
Greetings and
Resource
Salutations
Number of students
Welcoming
Inclusive language
Greeting
Teacher request
Closing remark
Colorado law
Faculty seeking feedback
Empathy
Welcoming
Negotiation
Accommodation
Contact information
Teaching / Facilitation
Questioning
Answering questions
Elaboration /
clarification
Positive feedback
Resource
Step 5. Identify themes that emerge from the data (include specific
language from the groups, codes, or data when appropriate)
While RTEOF have to deal with day to day course logistics, such as directions on how to complete
assignments and course expectations, they play more of a role of as a facilitator through the use of
questioning, elaborating/clarifying, and giving positive feedback than as a instructor or giver of
knowledge.
Figure 3.1. Steps followed to complete constant comparison analysis of online
discussions
Reliability and Validity
Reliability and validity are key considerations for any researcher. These two
concepts are intricately connected (Cresswell, 2008). Issues of reliability and validity are
addressed in the following pages.
Reliability
Reliability is essentially the consistency of scores researchers obtain from a
measure (Goodwin, 2001). More specifically, according to Goodwin, “interrater
agreement and reliability is the extent to which scores obtained from two or more raters

95. 80
(scorers, judges, observers) are consistent” (p. 15). The most common method used to
calculate interrater reliability is a percent agreement statistic (Rourke et al., 2001b).
I selected ten percent of the discussions to assess interrater reliability. Two coders
(i.e., me and another researcher) chunked and coded the discussions using content
analysis. One challenge though with interrater reliability is that there is not a consistent
agreed upon level of what must be achieved (Rourke, Anderson, Garrison, & Archer,
2001b). Past research on social presence (e.g., Rourke et al., 2001a; Swan, 2003) was
used as a guide of where interrater reliability should lie.
Following Rourke et al. and Swan, the entire discussion posting was used as the
unit of analysis. As a result, 100% agreement was found between the two coders when
identifying the chunks to code because the learning management system clearly identified
an entire post. After some initial training, the two raters then coded 10% of the
discussions to establish reliability of coding. A percent agreement statistic was calculated
using Holsti’s (1969) coefficient of reliability for each of the threaded discussions:
• Reading Group G: 80%
• Pair 6: 78%
• Project Group 3: 77%
The overall percent agreement for all of the discussion was 78%, which is an acceptable
level of agreement given past research (Garrison, Anderson, & Archer, 2001; Hughes et
al., 2007).
Validity
Validity is a complex concept. Validity has been defined as the “trustworthiness
of inferences drawn from data” (Eisenhart & Howe, 1992, p. 644). However, over the

96. years, researchers’ understanding of validity—and therefore definitions and standards—
81
has evolved (Dellinger & Leech, 2007; Goodwin & Leech, 2003). Further, quantitative
and qualitative researchers tend to understand and deal with validity differently
(Cresswell & Plano Clark, 2007; Dellinger & Leech, 2007).
Historically, quantitative researchers separate validity into content, criterion-related,
and construct validity. Qualitative researchers, on the other hand, historically
describe validity as trustworthiness. A large component of establishing trustworthiness is
developing a sound theoretical framework (Garrison, Cleveland-Innes, Koole, &
Kappelman, 2006, p. 2), as I have tried to do throughout this study. Further, the coding
schemes that were used for this study are based directly in the literature.
Chapter Summary
Researchers have studied social presence in online learning environments for a
number of years now. However, to date, research on social presence suffers from a host
of problems—ranging from inconsistent and contradictory findings to strange sampling
decisions. Further, researchers have not been able to demonstrate a consistent
relationship between student learning and social presence. Part of the problem might be
the methodological decisions that researchers have made. Rather than employ a mono-method
approach like the majority of past research, this study employed a mixed methods
approach to studying social presence—utilizing both quantitative and qualitative methods
to understand the complex nature of social presence. In addition, in this study I
specifically focused on how students establish and maintain social presence in a text-based
environment by focusing on what is “said” in the threaded discussions.

97. 82
CHAPTER 4
RESULTS
As described in Chapter 3, I used a mixed methods research approach to explore
how social presence manifests in an online graduate level education course at the
University of Colorado Denver. More specifically, I was interested in finding out how
users established their social presence through text alone in asynchronous threaded
discussions. In this chapter, I share the results from the word count, content analysis, and
constant comparison analysis I conducted to explore how online learners establish and
maintain their presence in one specific fully online course.
Word Count
I conducted a word count of the threaded discussions to initially explore the data.
I was curious whether certain types of words appeared more frequently than others across
all of the threaded discussions as well as within certain types of threaded discussions as
opposed to others. For example, did certain words appear more in threaded discussions
with a pair of students vs. threaded discussions with small groups of students? Using
Nvivo 8, I specified the parameters for a word count frequency report.
I first looked at the top 50 words used across all of the threaded discussions (see
Appendix B for a complete list of word count frequency’s for each word count
conducted). While I set forth to investigate the top 50 words used across all threaded
discussions, I found that the top 20 words were sufficient to get a basic understanding of
the data. Thus, I only report on the top 20 words in this section (though I have included
the top 50 results in Appendix B).

98. 83
Table 4.1 lists the top 20 words used across all of the threaded discussions (see
Figure 4.1 for a visual representation). The word “I” was used most frequently (4,858
times which represents 4.13% of all the words used) followed next by the word “you”
(2,186 times; 1.86% of all the words used). The frequency of these words is not that
surprising but the fact that “we” (which is often used as a sign of group reference and
therefore an indicator of social presence) was used 1,367 times (or 1.16% of all words
used) and ranks fourth overall in all words used is noteworthy. Some other things of
interest are the fact that “your” which can often be an example of “acknowledgement”
(i.e., another indicator of social presence) was used 810 times or eighth overall. And
finally, the word “policy”—which is the focus of the course—was used 600 times (or 10th
overall) whereas the instructor’s pseudonym, “Bob,” was used 566 times (or 14th overall).
Table 4.1 Top 20 Words Used Across All Threaded Discussions
Rank Word Count Percentage (%)
1 I 4858 4.13
2 you 2186 1.86
3 have 1428 1.21
4 we 1367 1.16
5 my 1001 0.85
6 what 948 0.81
7 do 814 0.69
8 your 810 0.69
9 can 730 0.62
10 policy 600 0.51
11 me 595 0.51
12 all 592 0.50
13 about 574 0.49
14 bob 566 0.48
15 so 565 0.48
16 Instructor 564 0.48
17 Think 553 0.47
18 Our 538 0.46
19 Work 494 0.42
20 Would 482 0.41

99. 84
Figure 4.1. Word cloud of word count results without the discussion headings.
After looking at the frequency of the top 50 words used across all threaded
discussions, I then ran a word count report for each of the main threaded discussions:
Project Groups (see Table 4.2), Pairs (see Table 4.3), and Reading Groups (see Table
4.4). While “I” and “you” were still the first and second most used words in each of the
main threaded discussions, Figure 4.1 illustrates that “we” and “your” (two possible
social presence indicators) show up in the top 20 across all three of these threaded
discussions and “our” (which is also a possible social presence indicator) shows up across
two of the threaded discussions—namely, the Project Groups and the Pairs threaded
discussions. Each of these words according to the coding sheet (which was discussed in
Chapter 3) and the literature in general (see Chapter 2) are possible indicators of group
reference and acknowledgement and therefore considered to be indicators of social
presence. I mention “possible” because in this case word count does not take into
consideration the context in which a given word is used; for instance, “we” could be
referring to “we Americans” or “we the class.”

100. 85
However, I found it interesting, though not necessarily surprising, that the word
“we” and to a smaller degree “our” (i.e., group reference) show up more in specific types
of small-group discussions where the purpose of the discussion is on collaborating on a
class project together as compared to reading groups (which are also small-group
threaded discussions but one with a different purpose). This suggests that the purpose of a
threaded discussion might influence the degree to which its participants employ certain
types of behaviors (i.e., the things referred to later in this chapter as indicators of social
presence) to establish and maintain their social presence. Once I had a basic feel for the
data, I then conducted a content analysis, which I elaborate on in the next section.
Project Groups Pairs Reading Groups
Rank Word Count % Word Count % Word Count %
1 I 1674 4.08 I 960 4.87 I 1784 3.79
2 you 729 1.78 you 438 2.22 you 802 1.70
3 we 678 1.65 my 339 1.72 have 532 1.13
4 have 497 1.21 have 291 1.48 we 416 0.88
5 what 387 0.94 we 209 1.06 what 358 0.76
6 do 267 0.65 your 189 0.96 do 348 0.74
7 can 261 0.64 me 148 0.75 policy 344 0.73
8 your 258 0.63 what 145 0.74 my 328 0.70
9 our 251 0.61 do 130 0.66 can 297 0.63
10 all 241 0.59 work 123 0.62 reading 293 0.62
11 think 221 0.54 about 119 0.60 your 283 0.60
12 my 215 0.52 goals 116 0.59 one 255 0.54
13 so 214 0.52 can 110 0.56 about 242 0.51
14 data 202 0.49 our 102 0.52 all 231 0.49
15 policy 193 0.47 how 100 0.51 think 227 0.48
16 would 184 0.45 school 97 0.49 me 226 0.48
17 some 181 0.44 teachers 85 0.43 so 225 0.48
18 need 176 0.43 goal 84 0.43 instructor 222 0.47
19 me 173 0.42 some 82 0.42 bob 221 0.47
20 about 171 0.42 would 82 0.42 summary 196 0.42
Figure 4.2. Frequency of possible social presence indicators across the three major
and most frequented threaded discussions.

101. 86
Table 4.2 Top 20 Words across Project Groups
Rank Word Count Percentage (%)
1 I 1674 4.08
2 you 729 1.78
3 we 678 1.65
4 have 497 1.21
5 what 387 0.94
6 do 267 0.65
7 can 261 0.64
8 your 258 0.63
9 our 251 0.61
10 all 241 0.59
11 think 221 0.54
12 my 215 0.52
13 so 214 0.52
14 data 202 0.49
15 policy 193 0.47
16 would 184 0.45
17 some 181 0.44
18 need 176 0.43
19 me 173 0.42
20 about 171 0.42
Table 4.3 Top 20 Words across Pairs
Rank Word Count Percentage (%)
1 I 960 4.87
2 you 438 2.22
3 my 339 1.72
4 have 291 1.48
5 we 209 1.06
6 your 189 0.96
7 me 148 0.75
8 what 145 0.74
9 do 130 0.66
10 work 123 0.62
11 about 119 0.60
12 goals 116 0.59
13 can 110 0.56
14 our 102 0.52
15 how 100 0.51
16 school 97 0.49
17 teachers 85 0.43
18 goal 84 0.43
19 some 82 0.42
20 would 82 0.42

102. 87
Table 4.4 Top 20 Words across Reading Groups
Rank Word Count Percentage (%)
1 I 1784 3.79
2 you 802 1.70
3 have 532 1.13
4 we 416 0.88
5 what 358 0.76
6 do 348 0.74
7 policy 344 0.73
8 my 328 0.70
9 can 297 0.63
10 reading 293 0.62
11 your 283 0.60
12 one 255 0.54
13 about 242 0.51
14 all 231 0.49
15 think 227 0.48
16 me 226 0.48
17 so 225 0.48
18 instructor 222 0.47
19 bob 221 0.47
20 summary 196 0.42
Content Analysis
After conducting word count, I used an amended version (see Chapter 3) of the
social presence indicators developed by Rourke et al. (2001a) to conduct content analysis
on all of the threaded discussions in the course in order to identify what types of social
presence indicators were present in each threaded discussion. As an exploratory study, I
was interested in exploring the data to see how the students and the instructor in this
given sample established and maintained their social presence. More specifically, though,
I was curious about the overall occurrence of all of the social presence indicators (taken
as a whole) across all of the threaded discussions, as well as the degree to which each
category (i.e., groups of specific types of social presence indicators) and specifically each

103. 88
individual indicator were used in this sample. But at the same time, based on the CoI
framework coupled with the word count results, I was also interested in the degree to
which all of the social presence indicators, categories of social presence indicators, and
specifically each individual social presence indicator occurred in specific types of
threaded discussion. Finally, and based in part on the results of my own research
(Lowenthal & Dunlap, 2011), I was curious how individual students might employ
certain types of social presence behaviors differently than others.
In summary, in order to explore how social presence manifests in threaded
discussions (i.e., the research question guiding this study), I was interested in the
occurrence and the frequency of the social presence indicators across all of the threaded
discussions, as well as their occurrence and frequency within specific threaded
discussions, and finally their relationship to each student (i.e., how often each student
used specific social presence indicators). Figure 4.2 visually illustrates the stages of
disaggregation I went through and report on in the following sections.
Figure 4.3. Stages of disaggregation of content analysis used to explore use of social
presence indicators in a fully online asynchronous online course.

104. 89
Stage One: Social Presence Categories and Indicators across All Threaded
Discussions
Past research on social presence, or at least research focused on identifying
indicators of social presence in online discussions, has focused primarily on reporting the
results in terms of the three categories or types of social presence indicators. Thus, I was
first interested in identifying which category of social presence indicators (i.e., Affective,
Interactive, and Cohesive) was identified the most and which was identified the least
across all of the threaded discussions. In other words, as a class, were “Affective,”
“Cohesive,” or “Interactive” indicators used the most?
Content analysis revealed that of the three different categories (or types) of social
presence, “Interactive” indicators were present the most with 2,581 instances, “Cohesive”
indicators were present the second most with 2,454 instances, and “Affective” indicators
were present the least with 1,373 instances (see Figure 4.4 and Table 4.5). The
differences between “Interactive” indicators and “Cohesive” indicators across all of the
threaded discussions are minor. But there is an observable difference between these two
categories and the “Affective” category of social presence indicators (see Figure 4.4). In
other words, in this sample, students used “Affective” indicators the least. This is
interesting in part because while Hughes et al. (2007) found a similar result in their
sample, Swan (2003) found that “Affective” indicators were actually used the most in her
sample.

105. 90
Figure 4.4. A visual depiction of the frequency of each of the three social presence
categories.
After examining the category level, I drilled down further to identify the
frequency at which participants in this sample used each of the individual social presence
indicators across all of the threaded discussions. The top three indicators used across all
of the threaded discussions were “acknowledgement” (i.e., recognizing and openly
acknowledging a previous post by a person) which was used the most at 1,137 instances,
followed next by “invitation” (e.g., asking a question) which was used 747 times, and
then by “vocatives” (i.e., addressing someone directly by the first name) which was used
748 times (see Table 4.5 and Figure 4.5).
It is difficult to compare these results to other researchers because as mentioned
earlier, the majority of those who do analyze online discussions focusing on social
presence indicators do not report their results at the indicator level. Swan (2003),
however, is one exception. But Swan only reports her findings at the indicator level
through a series of bar graphs that lack exact numerical values (but still enable a reader to
compare the frequency of each indicator). Acknowledgement was the only top-three

106. 91
indicator shared with my sample and Swan’s sample; paralanguage which was used
infrequently in this sample was actually the most frequently occurring social presence
indicator in Swan’s study.
Table 4.5 Social Presence Frequency across All Forums
Category & Indicator Frequency
Total Affective Responses 1373
Paralanguage (PL) 270
Emotion (EM) 526
Humor (H) 53
Self-Disclosure (SD) 524
Total Interactive Responses
2581
Acknowledgement (AK) 1137
Agreement / Disagreement (AG) 192
Invitation (I) 747
Expressing Appreciation (EA) 505
Total Cohesive Responses
2454
Greetings & Salutations / Phatics (GS) 714
Vocatives (V) 748
Group Reference / inclusivity (GR) 638
Embracing the Group (EG) 354
Total 6408
The least frequently used indicators of social presence were “humor” which was
used the least at 53 instances (which was also the least used indicator in Swan’s sample),
followed next by “Agreement/Disagreement” which was used 192 times, and then by
“paralanguage” which was used 270 times (see Table 4.6 for a complete ranking of each
of the social presence indicators across all of the threaded discussions).

107. 92
270
526
53
524
1137
192
747
505
714 748
638
354
1200
1000
800
600
400
200
0
Affective
Interactive
Cohesion
Figure 4.5. Frequency of social presence indicators across all threaded discussions
Table 4.6 Social Presence Indicators Ranking from Highest to Lowest Frequency
Social Presence Indicators Frequency
Acknowledgement 1173
Vocatives 748
Invitation 747
Greetings & Salutations / Phatics 714
Group Reference / Inclusivity 638
Emotion 526
Self-Disclosure 524
Expressing Appreciation 505
Embracing the Group 354
Paralanguage 270
Agreement / Disagreement 192
Humor 53
While it is useful to compare how the individual social presence indicators
manifest across all three categories of social presence, it is also helpful to drill down to
see how they compare to other indicators within their same category. The reason for this
is because it is possible that within a given category that certain indicators of social
presence are used more frequently than others. For instance, in the “Affective” category

108. 93
“emotion” and “self-disclosure” appeared the most frequently and almost in the same
frequency (see Figure 4.6). In the “Interactive” category however, signs of
“acknowledgement” were by far the most frequently used social presence indicator (see
Figure 4.6). Finally, in the “Cohesion” category, “greetings / salutations / phatics,”
“vocatives,” and then “group reference” all appeared in about the same frequency but
“embracing the group” was used the least (see Figure 4.6).
Figure 4.6. Social presence indicators separated by category

109. 94
Stage Two: Social Presence Categories and Indicators By Discussion Forum
As helpful as it is to look at the frequency of social presence indicators across all
of the threaded discussions and treating all of the threaded discussions essentially as one
case, it is perhaps more insightful and helpful to drill down and look at the occurrence of
social presence indicators across and within types of threaded discussions. At this stage, I
first looked at the occurrence of social presence indicators across specific types of
threaded discussions.
For the ease of reporting, I separated full-class threaded discussions (i.e.,
discussions that are “open” to the entire class) from small-group threaded discussions
(i.e., discussions that are “closed” to a small select group of students assigned with a
specific task like discussing the reading or collaborating on a course project). See Table
4.7 for the list of “open” vs. “closed” threaded discussions. But because each threaded
discussion differs in total number of posts and words, I needed a way to calculate the
social presence density of each discussion.
Following the lead of Rourke et al. (2001a), I calculated the social presence
density for each indicator in each threaded discussion. But because the unit of analysis
for this study was the entire post, I calculated the social presence density by taking the
average social presence indicator per post (as opposed to per word like Rourke et al.,
2001a) to facilitate comparison across open and closed threaded discussions.

110. 95
Table 4.7 Open vs. Closed Threaded Discussions
Open to Entire Class Small Group (limited to 2-5)
Virtual Office
--Virtual Office
General Discussions
--General Syllabus
--General Groups
--General Independent Work
--General Individual Work
Adult Learning Discussions
--Adult Learning Discussion Forum:
Your Learning
--Adult Learning Discussion Forum:
Questionnaire #1
Plus Delta Discussions
--Plus Delta Week 2
--Plus Delta Week 3
--Plus Delta Week 4
--Plus Delta Week 5
--Plus Delta Week 6
Reading Log Discussions
--Reading Log1
--Reading Log 3
Reading Groups
--Reading Group A
--Reading Group B
--Reading Group C
--Reading Group D
--Reading Group E
--Reading Group F
--Reading Group G
Pairs
--Pair 1
--Pair 2
--Pair 3
--Pair 4
--Pair 5
--Pair 6
--Pair 7
--Pair 8
--Pair 9
--Pair 10
Project Groups
--Project Group 1
--Project Group 2
--Project Group 3
--Project Group 4
--Project Group 5
I found when comparing the average social presence indicators per post between
open threaded discussions and closed threaded discussions that a higher density of social
presence occurred in closed threaded discussions than in open threaded discussions (see
Table 4.8). For instance, the average per post “Affective” indicator is 0.78 in closed
discussions (meaning there is an average .78 affective indicators per post) compared to
0.56 for open discussions; the average “Cohesive” indicators is 1.37 in closed discussions

111. as compared to 1.17 in open discussions; and the average “Interactive” indicators is 1.45
96
in closed discussions vs. 1.09 in open discussions.
Table 4.8 Average Social Presence Indicators Per Post across Open and Closed
Threaded Discussions
Open Discussions Closed Discussions
Total
Average
Total
Average
Affective 101 0.56 1272 0.78
Cohesive 211 1.17 2243 1.37
Interactive 197 1.09 2382 1.45
Total 509 2.81 5897 3.59
I then decided to look deeper within the closed discussions to explore any
observable differences between the different types of closed discussions used because
while all three were “closed” discussions, each one had a distinct purpose which could
have influenced how students posted in each threaded discussion. When comparing all
three of the different types of “closed” discussions (see Table 4.9 and Figure 4.7), “Pairs”
had the highest total social presence average per post with 4.20 social presence indicators
per post. “Project Groups” was next with an average of 3.76 social presence indicators
per post. And then “Reading” groups had the lowest average of social presence indicators
per post. These differences could likely be due to a combination of the group size and the
purpose of each of these threaded discussions. For instance, the “Pairs” and the “Project
Groups” had very specific tasks that required interaction, cohesion, and collaboration
whereas the “Reading Groups” (while also a small group) had less prescriptive tasks (see
Table 3.1 in Chapter 3).

112. 97
Table 4.9 Average Social Presence Indicators across Closed Threaded Discussions
Reading Groups Pairs
Project Groups
Total
Average
Total
Average
Total
Average
Affective 549 0.77 253 0.95 470 0.71
Cohesive 776 1.09 467 1.76 1000 1.50
Interactive 956 1.35 394 1.49 1032 1.55
Total 2281 3.21 1114 4.20 2502 3.76
Figure 4.7. Visual depiction of the average social presence indicators grouped by
category in closed threaded discussions.
But when I began to compare each category and later each indicator, the results
began to change. For instance, the “Pairs” threaded discussions have the highest average
of all of the social presence indicators per post across all the categories and indicators.

113. But when I disaggregated these results, I found that the “Pairs” threaded discussions did
not have the highest social presence density across all three categories of social presence.
For the interactive category of indicators, the “Pairs” group actually had a lower per post
98
average than the “Project Groups.” At the same time, while the “Reading Groups” had
the lowest total social presence average per post overall, these threaded discussions
actually had a higher average of affective indicators than “Project Groups” (see Table
4.10). This could suggest that certain types of tasks in certain group sizes could elicit
more social presence behaviors per participant than others. At the same time, the
differences are minor and more research would likely need to be conducted to support
this theory.
Table 4.10 Ranking of Average Social Presence Indicators Across Closed Threaded
Discussions
Social Presence Category & Closed Threaded Discussion Average Per Post
Affective Indicators
Pairs 0.95
Reading Groups 0.77
Project Groups 0.71
Cohesive Indicators
Pairs 1.76
Project Groups 1.50
Reading Groups 1.09
Interactive Indicators
Project Groups 1.55
Pairs 1.49
Reading Groups 1.35
Each threaded discussion—specifically the closed threaded discussions—consists
of different students and therefore even though the tasks might be the same, it is possible
that individual students and their natural or learned communication skills influence the

114. 99
frequency and therefore overall social presence density in a given threaded discussion
(which is in part why I looked at each student’s social presence behaviors during Stage 3
of the content analysis). Therefore, I dug deeper to look at the social presence density
across all closed threaded discussions (see Table 4.11).
One of the Pairs threaded discussions—specifically Pair 9—had the highest
overall average of social presence indicators per post per discussion as well as the highest
per post average of each of the three categories of social presence indicators. Reading
Group E and Reading Group G ended up with the lowest social presence per post average
per individual threaded discussions. These results follow the general trend identified
earlier (see Figure 4.6) with the Pairs threaded discussions having the overall highest
density of social presence per post and the Reading Groups threaded discussions having
the lowest overall density of social presence per post. This could suggest that the overall
size and purpose of a specific discussion highly influences the amount of social presence
indicators used by students in any given discussions: For instance, the Pairs discussions
involved two students taking part in personal discussions versus the Reading Groups
which involved small groups of 4-5 students talking about the weekly readings in the
course. As one might imagine, two students discussing personal matters might engender
more affective, cohesive, and interactive indicators than a larger group discussing course
readings.

115. 100
Table 4.11 Average Social Presence Indicator per Threaded Discussion
Discussion Forum Total
Posts
Affective/
Avg. Per
Post
Cohesive/
Avg. Per
Post
Interactive
/Avg. Per
Post
Social
Presence/
Avg. Per
Post
Open Discussions
Virtual Office 44 16 (0.36) 59 (1.34) 44 (1.00) 119 (2.7)
General: Syllabus 48 12 (0.25) 44 (0.92) 34 (0.71) 90 (1.88)
General: Groups 14 8 (0.57) 12 (0.86) 16 (1.14) 36 (2.57)
General:
3 3 (1.00) 5 (1.67) 3 (1.00) 11 (3.67)
Independent Work
General:
Individual Work
2 0 (0.00) 3 (1.50) 2 (1.00) 5 (2.5)
Adult Learning
Discussion Forum
–Your Learning
12 4 (0.33) 13 (1.08) 12 (1.00) 29 (2.42)
Adult Learning
Discussion Forum
–Questionnaire #1
3 4 (1.33) 2 (0.67) 5 (1.67) 11 (3.67)
Plus Delta Week2 13 15 (1.15) 24 (1.85) 15 (1.15) 54 (4.15)
Plus Delta Week 3 22 19 (0.86) 30 (1.36) 36 (1.64) 85 (3.86)
Plus Delta Week 4 2 3 (1.50) 0 (0.00) 3 (1.50) 6 (3.00)
Plus Delta Week 5 2 3 (1.50) 5 (2.50) 2 (1.00) 10 (5.00)
Plus Delta Week 6 3 7 (2.33) 4 (1.33) 4 (1.33) 15 (5.00)
Reading Log 1 12 7 (0.58) 10 (0.83) 20 (1.67) 37 (3.08)
Reading Log 3 1 0 (0.00) 0 (0.00) 1 (1.00) 1 (1.00)
Closed Discussions
Reading Group A 125 110 (0.88) 128 (1.02) 192 (1.54) 430 (3.44)
Reading Group B 132 88 (0.67) 124 (0.94) 203 (1.54) 415 (3.14)
Reading Group C 95 104 (1.09) 129 (1.36) 95 (1.00) 328 (3.45)
Reading Group D 109 120 (1.10) 153 (1.40) 186 (1.71) 459 (4.21)
Reading Group E 40 23 (0.58) 29 (0.73) 41 (1.03) 93 (2.33)
Reading Group F 106 59 (0.56) 84 (0.79) 126 (1.19) 269 (2.54)
Reading Group G 103 45 (0.44) 129 (1.25) 113 (1.10) 287 (2.79)
Pair 1 32 18 (0.56) 46 (1.44) 51 (1.59) 115 (3.59)
Pair 2 40 41 (1.03) 71 (1.78) 59 (1.48) 171 (4.28)
Pair 3 45 41 (0.91) 84 (1.87) 78 (1.73) 203 (4.51)
Pair 4+ 6 5 (0.83) 5 (0.83) 5 (0.83) 15 (2.50)
Pair 5 30 38 (1.27) 65 (2.17) 38 (1.27) 141 (4.70)
Pair 6 28 14 (0.50) 38 (1.36) 30 (1.07) 82 (2.93)
Pair 7 26 23 (0.88) 41 (1.58) 40 (1.54) 104 (4.00)
Pair 8 21 33 (1.57) 48 (2.29) 33 (1.57) 114 (5.43)
Pair 9 15 25 (1.67) 38 (2.53) 30 (2.00) 93 (6.20)
Pair 10 22 15 (0.68) 31 (1.41) 30 (1.36) 76 (3.45)

116. 101
Table 4.11 (con’t.)
Closed Discussions
(con’t.)
Project Group 1 109 72 (0.66) 160 (1.47) 167 (1.53) 399 (3.66)
Project Group 2 180 96 (0.53) 276 (1.53) 292 (1.62) 664 (3.69)
Project Group 3 138 111 (0.80) 168 (1.22) 189 (1.37) 468 (3.39)
Project Group 4 113 79 (0.70) 136 (1.20) 141 (1.25) 356 (3.15)
Project Group 5 126 112 (0.89) 260 (2.06) 243 (1.93) 615 (4.88)
Stage Three: Social Presence Categories and Indicators By Students
While conducting the content analysis, I began to get a sense that certain students
used certain social presence indicators (e.g., paralanguage and vocatives) more than
others. Therefore, I decided to investigate the frequency at which each student used social
presence indicators. I reasoned that it could be that, even though a certain threaded
discussion (which consisted of a group of students) might have a high social presence
density, it could be the result of one group member who was extremely active and
proficient with employing affective, interactive, and cohesive means of communication in
threaded discussions.
Henceforth, I first looked at each participant’s use of all three of these categories
of social presence as a whole; however, I excluded five students who failed to post more
than ten overall posts throughout the semester. Of those who posted more than ten times,
Cathy had the highest average with 5.43 instances of social presence per post, followed
next by Diana with 4.87 per post, and Mary with 4.64 per post. This becomes more
striking when these results are compared to participants with the lowest use of social
presence indicators per post. The three participants with the lowest number of social
presence indicators per post were Instructor Bob who had the lowest average at 2.24
instances per post, followed by Sam with 2.42 per post, and then Monica at 2.89 per post.

117. 102
But when I dug a little deeper I found that a high or low social presence rating
(i.e., the average social presence indicators used per post) does not necessarily mean that
the participant in question scores the same on all three categories of indicators or even on
a given set of indicators within a category. In other words, one could feel competent and
comfortable with interactive types of communication but not with affective or cohesive.
For instance, while Cathy had a high overall social presence average per post (when
taking into consideration all three categories of social presence), she had one of the three
lowest interactive averages per post. In other words, while her use of affective and
cohesive indicators was high compared to her peers, her use of interactive indicators was
low compared to her peers. Similarly, while Instructor Bob had an overall low total social
presence score, he in fact had the highest interactive score (see Table 4.12) thus
suggesting that he may be more proficient at interactive types of communication than
cohesive or affective.
Table 4.12 Student’s Use of Social Presence Categories
Total
Posts
Social
Presence
Total Posts
(Avg. Per
Post)
Affective
Total Posts
(Avg. Per
Post)
Cohesive
Total Posts
(Avg. Per
Posts)
Interactive
Total Posts
(Avg. Per
Posts)
Adam 76 254 (3.34) 56 (0.22) 109 (0.43) 89 (0.35)
Cathy 77 418 (5.43) 122 (0.29) 175 (0.42) 121 (0.29)
Christine 107 362 (3.38) 86 (0.24) 115 (0.32) 161 (0.44)
Daphne 73 253 (3.47) 42 (0.17) 112 (0.44) 99 (0.39)
Dawn 121 360 (2.98) 69 (0.19) 123 (0.34) 168 (0.47)
Denise 103 393 (3.82) 61 (0.16) 178 (0.45) 154 (0.39)
Diana 94 458 (4.87) 156 (0.34) 151 (0.33) 151 (0.33)
Erica 66 221 (3.35) 53 (0.24) 101 (0.46) 67 (0.30)
Gabriela 55 173 (3.15) 34 (0.20) 66 (0.38) 73 (0.42)
Instructor Bob 328 736 (2.24) 115 (0.16) 204 (0.28) 417 (0.57)

118. 103
Table 4.12 (con’t.)
Kate 99 354 (3.58) 52 (0.15) 157 (0.44) 145 (0.41)
Kyleigh 85 274 (3.22) 75 (0.27) 99 (0.36) 100 (0.36)
Laura 39 172 (4.41) 44 (0.26) 73 (0.42) 55 (0.32)
Mary 117 543 (4.64) 91 (0.17) 231 (0.43) 221 (0.41)
Micky 93 423 (4.55) 96 (0.23) 174 (0.41) 153 (0.36)
Monica 53 153 (2.89) 32 (0.21) 61(0.40) 60 (0.39)
Richard 31 130 (4.19) 23 (0.18) 61 (0.47) 46 (0.35)
Sam 78 189 (2.42) 50 (0.26) 55 (0.29) 84 (0.44)
Sara 50 229 (4.58) 54 (0.24) 88 (0.38) 87 (0.38)
Vicky 64 234 (3.66) 47 (0.20) 82 (0.35) 105 (0.45)
But even treating each social presence indicator within a given category equally
could perhaps be hiding certain trends. For instance, it could be that certain people are
strong with certain indicators in a given category but not others (for instance, someone
might have a high Affective category but simply because he or she is really proficient at
disclosing personal information and sharing emotion, but not at using paralanguage and
humor).
So, I decided to take a look at the students with the highest overall social presence
average per post (see Figure 4.8 and Figure 4.9). While Cathy had the highest social
presence per post average at 5.43 instances per post, Cathy’s (like Mary’s) strength
appears to be “greetings and salutations.” Diana on the other hand uses “paralanguage”
more frequently than “greetings and salutations.” Diana though was one of the students in
the Pairs 9 threaded discussion which had the highest per post average of social presence
indicators; it is important to note that she was paired with Sara who was fourth on the
overall list with the highest average of social presence indicators.

119. These results likely suggest two things. First, that just because someone may be
104
proficient at employing a certain type or category of social presence behaviors (i.e.,
affective, interactive, and / or cohesive) does not mean that this same person is proficient
at or comfortable with each indicator related to the category of social presence
communication. In other words, while someone might use a lot of affective types of
communication, he or she might never use paralanguage and vocatives, opting instead for
the use of greetings and salutations, acknowledgement of others, and the use of emotion.
Second, these findings might point to the fact that people—especially in small groups—
might begin to mirror the communication behaviors of their peers. For example, if a peer
(in a small group) has strong social presence behaviors and heavily uses paralanguage
then other students in the group might begin to use paralanguage more frequently than
before simply from mimicking their peer.
Cathy Diana Mary
Greetings &
salutations
0.84 Paralanguage 0.64 Greetings &
salutations
0.85
Emotion 0.6 Acknowledgement 0.63 Acknowledgement 0.79
Acknowledgement 0.6 Group Reference 0.62 Invitation 0.5
Vocatives 0.56 Invitation 0.59 Group Reference 0.5
Paralanguage 0.52 Emotion 0.5 Vocatives 0.44
Group Reference 0.51 Self Disclosure 0.49 Expressing
Appreciation
0.44
Invitation 0.45 Greetings &
salutations
0.43 Emotion 0.37
Expressing
Appreciation
0.42 Vocatives 0.31 Self Disclosure 0.31
Embracing the
Group
0.36 Expressing
Appreciation
0.31 Embracing the
Group
0.18
Self Disclosure 0.35 Embracing the
Group
0.26 Agreement 0.17
Humor 0.12 Agreement 0.09 Paralanguage 0.1
Agreement 0.1 Humor 0.03 Humor 0
Figure 4.8. Ranking of social presence indicators used by the three students with the
highest overall social presence per post average.

120. 105
Mary’s Individual Use of Social Presence Indicators
0.1
0.37
0
0.31
0.85
0.44
0.5
0.18
Diana’s Individual Use of Social Presence Indicators
Cathy’s Individual Use of Social Presence Indicators
0.5
0.79
0.17
0.44
Paralanguage
EmoSon
Humor
Self Disclosure
GreeSngs & salutaSons
VocaSves
Group Reference
Embracing the Group
Acknowledgement
InvitaSon
0.64
0.5
0.03
0.49
0.43
0.62 0.31
0.26
0.63
0.59
0.09
0.31
Paralanguage
EmoSon
Humor
Self Disclosure
GreeSngs & salutaSons
VocaSves
Group Reference
Embracing the Group
Acknowledgement
InvitaSon

121. 106
0.52
0.6
0.12
0.35
0.84
0.56
0.45
0.1
0.51
0.6
0.36
0.42
Paralanguage
EmoSon
Humor
Self Disclosure
GreeSngs & salutaSons
VocaSves
Group Reference
Embracing the Group
Acknowledgement
InvitaSon
Figure 4.9. Disaggregation of three students with highest social presence per post
average.
Constant Comparison Analysis
While social science has a long tradition of using content analysis alone to
analyze the content of online discussions, I decided at the beginning of this study to use
multiple types of analysis in an effort to better explore how social presence manifests in
threaded discussions in a completely online course. I turned to constant comparison
analysis in hopes that it would reveal things that were missed by content analysis.
At the beginning of this study, I decided to analyze the threaded discussion with
the highest average social presence density per discussion post and the threaded
discussion with the lowest in hopes of identifying different ways that social presence
manifests in threaded discussions. After conducting content analysis, I selected the Pair 9
threaded discussion as having the highest social presence density at 6.20 per post and the
Reading Group E as having the lowest social presence density at 2.33 per post. I then
used constant comparison analysis to code these two threaded discussions in an effort to

122. see if themes might emerge that tell a similar or different story than the content analysis
107
results.
Due to the different nature of each threaded discussion, I conducted constant
comparison analysis on each discussion separately. I first analyzed Reading Group E
(which had the lowest social presence density). As touched on in Chapter 3, the Reading
Group discussions consisted of small groups of 4-5 students that were tasked with
discussing the course readings and jointly writing nine different reading logs about the
course readings. The readings logs were supposed to not only summarize the readings but
also bring up any questions the group members had so that the instructor could then
answer them in the Reading Group threaded discussions. Students had two incentives to
take part in the Reading Group threaded discussions: First, students were graded on each
of the nine reading logs, which consisted of 15.25% of the course grade; second, students
were graded for their online interactivity and quality of work, which consisted of 16.95%
of the course grade.
As mentioned in Chapter 3, the first step of conducting constant comparison
analysis involved reading the entire threaded discussion. After reading all of the posts in
the threaded discussion, I then chunked the text into meaningful units. I then coded each
meaningful unit while constantly comparing new codes with previous codes. During the
coding process, I focused on the way people were communicating while still trying not to
limit myself or be confined in any way to the social presence indicators used for content
analysis. After coding all of the meaningful units, I then listed the codes and grouped
similar codes.

123. The analysis resulted in 89 unique codes (see Appendix C). Those codes were gathered
108
into eight separate groups (see Table 4.13).
Table 4.13 Groups of Codes Resulting from the Constant Comparison Analysis
Reading Group E
Grouping of Codes
Course logistics & facilitation
Emotion
Greetings and Salutations
Sharing Life Details
Gracious/Gratitude
Self Disclosing Personal Matters
Playing Nice with Others
Policy Related Class Discussions
In the fifth and final step, I identified themes from the data—while including
specific language from the groups, codes, or data when appropriate. The following two
themes emerged from the data from Reading Group E. (I have italicized any text that
came straight from the threaded discussions.)
• Policy is complex and multifaceted; it is something that many students and
teachers have no idea about; while the readings varied in complexity and
required a little more time than texts in past classes, with the help of the
instructor the students came to find the study of policy interesting and
relevant.
• Students began the threaded discussion (which spanned two months) with chit
chatting and telling personal stories but quickly changed their focus to the task
at hand of discussing public policy in general and the readings in particular;

124. 109
overtime the focus of the discussion was solely on the reading and public
policy—by this point the discussion largely consisted of students posting
questions and the instructor answering the questions.
After analyzing the Reading Group E threaded discussion, I analyzed the Pair 9
threaded discussion (which had the highest social presence density) following the same
steps as above. The Pair 9 threaded discussion had a different purpose than the reading
group. According to the course syllabus, the Pairs group is a place where group members
work on a personal-professional development activity that requires each student to take a
bit of risk and develop some trust with each other while discussing individual personal-professional
goals that are deeply important to one another. Similar to the Reading
Group threaded discussions, students had two incentives to take part in the Pair’s
threaded discussions: First, students were graded on the 3-5 page paper that resulted from
their work in their pairs group they were assigned to, which consisted of 12.7% of the
course grade; second, students were graded for their online interactivity and quality of
work, which consisted of 16.95% of the course grade.
Likely due in part to the different purpose, the Pairs threaded discussions had a
higher social presence density than other threaded discussions but specifically the Pairs 9
group had the highest among all of the Pairs and all of the threaded discussions in
general. Analysis of the Pairs 9 group resulted in 63 codes (see Appendix C), which I
then grouped into nine groups (see Table 4.14).

125. 110
Table 4.14 Groups of Codes Resulting from the Constant Comparison Analysis
Reading Pair 9
Grouping of Codes
Course logistics & facilitation
Collaboration
Emotion
Sharing Life Details
Playing Nice with Others
Policy Related Class Discussions
Greetings and Salutations
Self Disclosing Personal Matters
Gracious/Gratitude
Three themes emerged from this data as well. Like before, I have italicized any text that
came straight from the threaded discussions.
• Students who have a past relationship and spend time with each other either
professionally (e.g., we are fortunate enough to work together) or personally
outside of class can have an easier time collaborating with each other because of
their past relationship, shared experiences, and geographic closeness which others
might not have. These benefits can help them NOT to be alone, give them
opportunities to chat a lot, provide a strong and safe foundation to openly share
how they are struggling personally and professionally, and to regularly meet face-to-
face.
• Instructors can only react to what they see in a threaded discussion. It is difficult
to assess and to support students when they collaborate offline.
• When asked to take a risk, trust a peer, and self-disclose personal details, it helps
when two people already know each other, have some trust already built, have
shared experiences, and finally have the ability to talk and meet offline.

126. 111
While the results of the constant comparison analysis did not necessarily
contradict any of the findings from the word count or content analysis, they did begin to
fill in some details about what students were talking about in each threaded discussion
and how the type and purpose of a threaded discussion could influence how people
communicate with one another.
Chapter Summary
I utilized word count, content analysis, and constant comparison analysis to
explore how social presence manifests in a fully online discussion. Results illustrate that
participants’ use of social presence behaviors (e.g., indicators of social presence) vary
across the course. The results also reveal that looking at the total social presence
indicators or even simply the frequency at which each category of social presence is used
(e.g., affective, cohesive, and interactive) might be misleading and miss important details
about how and when people use certain social presence behaviors. These results will be
discussed at greater length in Chapter 5.

127. 112
CHAPTER 5
DISCUSSION
I set out to explore how social presence manifests in a fully online asynchronous
course. In Chapter 1, I laid out an argument for why additional research needs to be
conducted on social presence. Then in Chapter 2, I reviewed the literature on social
presence and the community of inquiry (CoI). After reviewing the literature, I then
explained in Chapter 3 the methods that were used for this study. Finally in Chapter 4, I
reported on the results of the study. Now in Chapter 5, I will discuss the significance of
these results, the limitations of this study, and the practical implications for the results—
specifically for course designers and faculty.
Key Findings
A deep and meaningful educational experience involves teaching presence, social
presence, and cognitive presence (Garrison et al., 2000). The CoI framework posits that
social presence is developed as the result of teaching presence. More specifically,
educators develop social presence through instructional design and organization,
facilitating discourse, and direct instruction (the three components of teaching presence)
(Anderson, Rourke, Garrison, & Archer, 2001). This does not mean that social presence
cannot naturally occur. Walther (1992) argued almost 20 years ago that people are social
creatures and that given enough time people will find ways to use any communication
medium for social purposes. Online educators however typically do not want to wait and
hope that their student’s natural social tendencies kick-in. Instead they often strive to find
ways to help encourage the development of social presence in online courses (which is
what the CoI refers to as teaching presence).

128. 113
The CoI framework (as well as the CoI literature as whole), though, does not
provide much guidance on how to design courses, facilitate discourse, and provide direct
instruction to facilitate the development of social presence (Garrison & Arbaugh, 2007).
For instance, how many threaded discussions should be in a course? Should the threaded
discussions be full class discussions or small groups? Should they have specific
instructional tasks? Educators can make some inferences from the indicators of teaching
presence developed by Anderson et al. (2001) (see Table 5.1), but even these indicators
lack sufficient detail.
Table 5.1 Teaching Presence Categories and Indicators
Teaching Presence Categories and Indicators
Instructional Design and Organization
Setting Curriculum
Designing Methods
Establishing Time Parameters
Utilizing Medium Effectively
Establishing Netiquette
Facilitating discourse
Identifying areas of agreement/disagreement
Seeking to reach consensus/understanding
Encouraging, acknowledging, or reinforcing student contributions
Setting climate for learning
Drawing in participants, prompting discussion
Assess the efficacy of the process
Direct Instruction
Present content/questions
Focus the discussion on specific issues
Summarize the discussion
Confirm understanding through assessment and explanatory feedback
Diagnose misconceptions
Inject knowledge from diverse sources e.g., textbook, articles, internet, personal
experiences (includes pointers to resources)
Responding to technical concerns

129. Some of the results presented in Chapter 4 might begin filling this void. That is,
114
the results provide a couple of possible guidelines for how educators can design and
develop online courses to increase social presence. However, as an exploratory study
using a small sample, the findings from this study should not be generalized to all
populations. In fact, any and all findings should be confirmed with additional research.
With that in mind, I will address some key findings in the following paragraphs.
Group Size
One of the first things that stood out initially with the word count results and then
with the content analysis results was that the social presence density—that is, the average
social presence indicator per discussion post—differed across types of threaded
discussions, specifically open vs. closed discussions. In other words, a higher social
presence density existed for small-group discussions than for large-group discussions.
This suggests that students projected themselves as “real” and “there” in the threaded
discussions through specific social presence behaviors (e.g., self disclosing information,
addressing people by first name, using emoticons) more frequently in small discussions
than in large discussions.
While very little research has been conducted on group size and social presence,
Tu and McIsaac (2002) claimed that “appropriate communication group size” can
influence social interaction and thus social presence. They concluded based on the
qualitative data in their study that “the size of the discussion group exerted a major
impact on students’ interaction, particularly in real-time discussions” (p. 145). And while
they recommend that two to three participants are an ideal group size for real-time

130. 115
discussions, they unfortunately do not offer any suggestions for an ideal group size for
asynchronous discussions.
Rourke and Anderson (2002a) conducted a study on using peer teams to lead
discussions. They found that students preferred small-group peer-led threaded discussions
more than full class instructor-led discussions. They concluded that this preference was
possibly due to the fact the small-group discussions consisted of four students and were
led by their peers rather than the instructor. But the students’ preference for small-group
discussions could have been due to a combination of the group size, the instructional task,
and the instructor’s reduced role rather than simply the fact that the discussions were peer
led.
This finding about large- and small-group discussions, however, does not suggest
that social presence cannot develop in large group discussions. In fact, Nagel and Kotze
(2009) found high levels of social presence in a “super-sized” course of 100+ students.
This finding about group size might simply confirm what Kreijns, Kirschner, and
Jochems (2003) argued about group size—namely, that anonymity and non-participation
increases as groups get larger (p. 340). In other words, as the group size (or class size)
increases, it is easier for students to hide and sit back and lurk (or not participate at all).
Lurking is not necessarily a bad thing (see Dennen, 2008). However, students need to
actually interact with their peers in order to project themselves as “real” and “there” in
threaded discussions. And this type of interaction might simply be easier for students in
smaller groups—especially those who might have a tendency to lurk in large threaded
discussions.

131. My personal experience teaching in face-to-face environments has shown me that
as the class size gets larger, fewer and fewer students ask questions on their own. In my
experience, small groups can force even the shiest and reluctant student to talk to her or
116
his peers. Given this, it might make sense for faculty to utilize small-group threaded
discussions more at the beginning of a course to help students begin to establish their
social presence early on in a given course or program of study. Small groups likely place
an additional amount of peer pressure on individual students. Individual students are no
longer simply held accountable for their actions by their instructor but also by their peers.
In my experience, peers are much more likely to send other peers an email for
nonparticipation in small groups—especially those that involve group work—than they
would in large group discussions. Further, these results could support the need for the
development of a number of small learning communities rather than the typical approach,
which too often focuses on developing one all-encompassing learning community with
every student in the course. More research though needs to be conducted across other
samples to confirm that group size can in fact influence the development of social
presence.
Instructional Task
In this study, though, group size alone did not guarantee a high level of social
presence. For instance, project groups and pairs had a higher social presence density than
reading groups even though reading groups were also small groups. This difference in the
social presence density likely could be due to the instructional task of each threaded
discussion. In my experience, students quickly identify what discussions they need to
take part in and which one’s they do not—both in terms of the relevance of the threaded

132. discussions toward the course and the student’s personal and professional goals as well as
117
any points the discussion is worth toward the final grade.
Students’ participation in both of these threaded discussions was graded and both
of these threaded discussions were tied to specific assignments that were graded as well.
However, as mentioned in Chapter 4, the reading groups involved identifying questions
that resulted from the course readings and then having the instructor answer the
questions. As a result, the dynamic of the discussions appeared to be less goal specific (or
at least less clearly defined) as the other two types of small threaded discussions. Reading
Groups had less peer accountability at least in comparison to the Pairs threaded
discussion. Also, more student-to-instructor and instructor-to-student rather than student-to-
student interaction occurred in these threaded discussions. In fact, when looking at the
number of posts and the number of words in each post in these threaded discussions, the
instructor’s role in the reading groups is more prominent than in the Pairs or Project
groups (see Table 5.2). This does not necessarily mean that instructors should say less or
avoid direct instruction. In fact, the CoI argues for the use of direct instruction as one way
to establish social presence. Rather it might simply suggest that the purpose of a
discussion likely influences how and what a student posts—and therefore the amount of
social presence behaviors used by both faculty and students.
Table 5.2 Instructor vs. Student Postings in Small Discussions
Reading Groups Pairs Project Groups
Posts Words Posts Words Posts Words
Student 543 (77%) 42,176 (71%) 219 (83%) 22,087 (89%) 630 (94%) 47,758 (91%)
Instructor 165 (23%) 16,860 (29%) 46 (17%) 2,629 (11%) 42 (6%) 4,992 (9%)

133. 118
The pairs discussion groups had the highest overall density of social presence.
While this is likely due in part to the fact that the pairs groups consisted of only two
students, it is perhaps equally due to the fact that the pairs groups were tasked with
sharing personal things with one another. In fact, the pairs had the highest frequency of
affective indicators per post, which is likely largely due to the instructional task. To date
though, no research specifically examines how specific instructional tasks in threaded
discussions affect social-presence behaviors used in the threaded discussions.
Researchers for years have questioned how best to structure threaded discussions
(Gilbert & Dabbagh, 2005). And they have shown that the structure of a threaded
discussion as well as how an instructor posts—thus modeling and setting the tone—can
influence how students post (see Dennen, 2005). While Lowenthal and Dunlap (2011)
investigated students’ perceptions of how specific instructional tasks influence students’
perceptions of social presence, to date there is a lack of research on how small working
groups (working on specific assignments—whether group assignments or not) can help
build social presence.
The reason the pairs group had a higher social presence density though could also
be due in part to the instructors role in these discussions. An, Shin, and Lim (2009) found
that “when the instructor’s intervention was minimal, students tended to more freely
express their thoughts and opinions, with a large number of cues for social presence” (p.
749). Thus, while course designers like myself as well as faculty in general seem to
prefer clear-cut guidelines, it is possible that there are not any clear-cut guidelines. These
results seem to suggest that it could be a combination of small group size, instructional
tasks that engender interpersonal dialogue, and low instructor involvement that helps

134. build social presence. But at this point, while this is reasonable, it is simply speculation.
119
Additional variables such as one’s personal communication style, how discussions are
graded, and the relevance of the instructional tasks to name a few, need to be investigated
to see how they too influence how social presence manifests. Further research needs to
be conducted to verify how instructional tasks (including not only what students are
asked to do but also how they are graded as well as the personal and professional
relevance of the assignments), group size, and instructor involvement can impact the
development of social presence.
Past Relationships
Constant comparison analysis of the threaded discussions with the highest and the
lowest density of social presence revealed that the pairs with the highest social presence
density worked together and even carpooled together. Practitioners have argued for years
that online courses—whenever possible—should start with face-to-face meetings to
establish social presence. This finding, though, might suggest something more. It could
suggest that people who have a strong relationship outside of class might have an easier
time with interactive, cohesive, and affective types of communication than people who do
not have a relationship outside of class. This finding is supported by the work of
Lowenthal and Dunlap (2011). Lowenthal and Dunlap found that having a positive group
project experience with a student helps increase a student’s perceptions of social presence
between the students in question and helps them maintain future relationships with one
another—even in the absence of ever meeting face-to-face.
Both of these findings suggest that having a past relationship with someone is
helpful when establishing social presence in online courses. It could be that a cohort

135. 120
model that enables students multiple opportunities to build relationships with other
students semester after semester is more valuable (at least when it comes to building
social presence) than beginning a course or a program with face-to-face meetings.
Walther (1994) argued years ago that the possibility of future interaction can influence
the degree to which people socially interact online thus further giving support for cohort
models or other types of models that enable students to take multiple courses with the
same students and/or with the same instructor. Further research though is needed to
confirm this because while the students’ past relationship emerged in the data in this one
group, it was difficult to ascertain whether or not other students had past relationships
with their peers and if so to what degree.
One Size Doesn’t Fit All
But perhaps the number one finding of this study from a design perspective, as
disheartening as it is, is that one size does not fit all. In other words, the results show that
while there are trends (e.g., that closed threaded discussions had a higher social presence
density than open threaded discussions), there is not always a clear reason as to why some
students use specific social presence behaviors (e.g., paralanguage) and others do not.
While some students might use (or some threaded discussions might elicit) high levels of
social presence overall, each of the indicators or at least the categories (i.e., types of
social presence) differed across students and types of threaded discussions.
This finding supports what Lowenthal and Dunlap (2011) found. They found that
each student appears to have her or his own threshold for social presence. In other words,
different people have their own social presence needs. What works for one student might
not work for another and what is comfortable or ideal for one student might not be

136. 121
comfortable or ideal for another. Along these lines, it is possible that each person—
perhaps based in part on his or her own social presence needs—has developed their own
level of proficiency at utilizing social presence behaviors in threaded discussions. That is,
each person has developed different levels of literacy at computer-mediated discourse.
However, a stylistic element appears to affect how people communicate in online
learning environments as well. For instance, some students appear to almost habitually
use emoticons (like Diana) whereas others do not appear to use them at all (like Kate,
Denise, Dawn, or Laura). It is possible that just as people have different communication
styles in face-to-face environments, that they also have different communication styles in
online environments. Further research though is needed to find out why some people use
certain types of communication behaviors (e.g., the use of vocatives or paralanguage) and
others do not.
Limitations of Studying Social Presence
Every research study has some limitations. I address the limitations of this study
later in this chapter. For now, though, I want to address some insights that resulted from
studying social presence behaviors in threaded discussions in this study. These insights
are possible limitations of social presence theory in general (or at least how social
presence is conceptualized within the CoI) as well as possible limitations with identifying
and quantifying social presence behaviors in particular. While practitioners will likely
find little use of these insights, researchers of social presence on the other hand might
find this section the most useful contribution of this study.

137. 122
Situational Variables of CMC
As mentioned earlier, social presence theory dates back to the work of Short et al.
(1976). Short et al. developed their theory of social presence based on their research on
how telecommunications effects the way two people communicate. In other words, Short
et al. and their theory of social presence originally focused on one-to-one communication.
While instances of one-to-one CMC (e.g., email) occur, more often than not CMC in
online courses takes place in threaded discussions that involve three or more
communicators. Instances of one-to-one communication are found in threaded
discussions. But this one-to-one communication is often done “in front” of others. The
“publicness” of CMC in threaded discussions is likely to influence what, when, and how
a person communicates in online courses—which is perhaps why Tu focused so much on
privacy in his early work (see Chapter 2 and Tu 2000, 2001, 2002a, 2002b) and perhaps
why students feel more comfortable or more pressured to present themselves as “real”
and “there” in small personal groups as opposed to larger impersonal groups.
More often than not, though, communication in threaded discussions is a one-to-many
model—thus changing the dynamic and making it more like public speaking. Or
when it is one-to-one, it is like talking to another person on the phone but while on a
speakerphone (where others are listening). I contend that these changes in the social
context in which one communicates—more than any limitation of the technology—likely
changes how people communicate and establish themselves as “there” and “real.” This
becomes important when one starts to think about the indicators of social presence
developed by Rourke et al. (2001a).

138. 123
Rourke et al. (2001a) claim that they developed these categories and indicators
based on their previous work (Garrison, Anderson, & Archer, 2000), other literature in
the field, and finally their experience reading online transcripts. However, I posit that we
have reason to believe that, as the technology of online threaded discussion forums has
improved over the years, as bandwidth has increased, as people’s experience using CMC
has improved (e.g., the increase in email use and the fact that Facebook has millions of
users alone are great examples of how people’s use or at least comfort and ability with
CMC has improved), and finally as the pedagogies used in online courses have matured,
the study of online transcripts has or at least should have changed over the years. In other
words, many of these indicators of social presence might no longer be relevant, might
lack enough specificity, or simply might be based too much on old assumptions of so-called
“proper” ways to communicate with CMC (which were likely influenced by the
older one-to-one model of CMC).
For instance, while addressing someone by his or her first name might help build
a sense of closeness and presence, the genre of CMC that takes place in online courses—
especially in large threaded discussions—often makes it difficult to use certain types of
social presence behaviors like addressing someone by her or his first name. For instance,
when an instructor is addressing the class as a whole, it does not make sense to begin a
post by mentioning everyone’s name. Further, while it might make sense to begin an
initial reply to someone’s post by stating her or his first name (to build cohesion), as a
thread continues and the posts go back and forth, an insistence on beginning each post
with someone’s first name could influence the ebb and flow of a conversation and
possibly hurt cohesion by making the conversation feel overly formal. Another example

139. 124
of the issues that arise when beginning a post with someone’s first name is that times
occur when an instructor is responding to one student but wants to invite the entire class
into the discussion. If an instructor begins the reply with the student’s name, then it may
send a message to the other students that the post is only for that person and not the rest
of the class.
Another situational variable that is given very little attention in social presence
theory in general and specifically in the CoI framework (see Garrison et al., 2000) is how
one’s role or status can influence not only how but what one communicates and how one
is perceived as being “there” or being “real.” For instance, it is reasonable to assume that
students in an online course—even in a so called “learner centered” course—are more
interested in what their instructor has to say than their peers (if only because the
instructor will be assigning their grade at the end of the semester). In fact, eCollege—a
Learning Management System used at the University of Colorado Denver—recently
started highlighting instructors’ posts with a different color to differentiate them from the
rest—thus suggesting that instructors’ posts are different than students’ posts.
While the CoI framework has an element called “teaching presence,” as
mentioned earlier, it focuses on how instructors design and organize a course, facilitate
discourse, and provide direct instruction. Teaching presence, though, does not
specifically address how an instructor establishes his or her own social presence,
especially given the added task of direct instruction and facilitating discourse.
I have argued elsewhere (Lowenthal & Lowenthal, 2010), in part building on the
work of Swan and Shih (2005) and their differentiation between students’ and instructors’
social presence, that one problem with the CoI framework is that it does not differentiate

140. (or really even acknowledge) how an instructor might establish his or her social presence
125
differently than students. In my experience, instructors often talk differently than their
students—this happens both in face-to-face classrooms and online. Further, each
instructor has her or his own style and level of comfort in the classroom. While some
instructors share parts of their personality and will engage in affective types of
communication, others will not. Further, while instructors might build opportunities to
establish social presence in their own online courses—in my experience, they often will
not engage in these activities with students.
The bottom line is that when instructors talk (i.e., post), students tend to listen
(i.e., read). This is not always the case when other students talk. Students are not always
as interested in what their peers say as in what their instructor says. I often think about
what instructors do to establish their own social presence and how the little things they do
(because of their status) can carry even more weight than if a fellow student did the exact
same thing. For instance, I posit that, when an instructor engages in affective
communication (e.g., sharing emotion or self-disclosing), it carries more weight than
when a student does the same thing. Further, and because of the difference in roles and
status, students tend to talk to an instructor differently than to their peers (i.e., code
switch; see White & Lowenthal, 2011). But none of these dynamics are considered when
researchers study social presence.
Further, as the result of this study, I have begun thinking about ways that
instructors and students can actually thwart social presence. For instance, what happens
when a student posts a question that nobody acknowledges or responds to? While this
likely happens in most courses at least once a semester—if only because some students

141. post at the last minute in a given week—as a result of this study, I have begun thinking
about how detrimental it can be if a student self-discloses personal information or shares
126
emotional things and nobody responds or acknowledges it. Occurrences like this could
possibly result in students feeling alienated and not acknowledged as being “there” and
“real.” But so much of the literature on social presence focuses on what people do to
establish social presence rather than on things people can do to thwart social presence.
Short et al. (1976) originally studied types of communication that were not only
one-to-one but also ongoing in the given moment. I contend that asynchronous threaded
discussions in online courses that take place over time, involve a many-to-many model,
likely involve students who have past relationships with each other (e.g., from past
courses) and likely future relationships (e.g., future courses), and consist of individuals
who are most likely paying money to be involved in the threaded discussions (and
therefore have some extra motivation to effectively communicate with one another and
their instructor) are a bit more complicated than Short et al. and possibly even Rourke et
al. (2001a) could have originally imagined. I also contend that situational variables like
these need to be considered when studying social presence. For instance, while content
analysis is a useful technique to study online discussions, quantitative measures or counts
of social presence behaviors might have limited value—especially when they do not take
into consideration the context in which social behaviors are used.
Unit of Analysis
Among other things, the unit of analysis one uses when conducting content
analysis influences the frequency of social presence indicators. For instance, following
past researchers’ lead (e.g., Rourke et al., 2001a and Swan, 2003), for this study I used

142. the entire discussion post as the unit of analysis. While I do not regret this decision, I now
recognize that the unit of analysis one chooses can largely determine what one sees and
127
what one does not see in her or his findings.
I assert that when researchers approach analyzing online threaded discussions
from a purely quantitative content-analysis perspective—frequency counts are
everything. If researchers only count a specific indicator of social presence (e.g., use of
emotion) once in a post because the post is the unit of analysis, he or she is likely to miss
some details. For instance, you can imagine how many times students might use the word
“we” as a group reference within a single post in small-group discussions focused on a
group project. But if the unit of analysis is simply the entire post, the high frequency of
the use of the word “we” may be lost in the totality of the words. I posit that the
frequency of this group reference—the word “we”—would be captured more accurately
if the unit of analysis was smaller than the entire post (e.g., each meaningful unit). For
example, if a discussion post has the group reference “we” five times in the post, this
indicator of social presence would only be counted once if the unit of analysis is the
entire post but might be counted up to five times if the unit of analysis was a meaningful
unit (which is not always but often the sentence level).
Researchers have written much about the ideal unit of analysis when using
content analysis to code online discussions (De Wever, Schellens, Valcke, & Van Keer,
2006; Rourke & Anderson, 2004; Rourke, Anderson, Garrison, & Archer, 2001b).
Unfortunately, very little consensus exists on which is the best approach to take because
while one might gain granularity with using a smaller unit of analysis, interrater
reliability decreases and workload increases. I finally decided to stick with using the

143. 128
entire discussion post as my unit of analysis after hearing Wise (Wise & Chiu, 2011)
justify her decision for using the entire post as her unit of analysis at AERA. She argued
that students read and therefore interact with and make meaning from each post in
threaded discussions not with each paragraph or word. Future research must investigate
how the unit of analysis influences content analysis results of threaded discussions.
Problems with the Social Presence Indicators and Treating Them Equally
To truly understand social presence, researchers ideally should look at both
students’ attitudes of social presence as well as students’ behaviors online. In other
words, researchers need to get a better idea of what specific behaviors elicit perceptions
of “closeness” and “realness” in others. The indicators of social presence are a great start
but they have limitations (as touched on earlier). For instance, when using them to
conduct content analysis, a researcher is supposed to identify when they find an instance
of each indicator. Let’s take greetings and salutations as an example. The problem with
this is that greetings and salutations, while similar, are two different things. For instance,
one could argue that someone who continually uses a salutation more than a greeting is
focusing more on themselves than on acknowledging others in a given threaded
discussion. Further, a greeting with a vocative (e.g., “Hi John”) is arguably better at
developing a sense of presence and projecting oneself as “real” and “there” than either
“Hi” or ending a post with one’s first name.
Similarly, the current coding sheet lists paralanguage as a type of affective
communication that establishes social presence. The problem, though, with using
paralanguage as an indicator of social presence is two fold: First, all uses of paralanguage
are not necessarily equal; second, students use paralanguage differently. Regarding the

144. 129
first point, some students appear to be chronic users of ellipses and seem to almost use
them as a period or a pause rather than in an emotive sense. This use of paralanguage is
different than intentional uses of emoticons and should arguably be treated as being
different. Secondly, paralanguage—especially the use of anything more than
emoticons—seems to be a learned behavior that only certain types of students use. In
other words, if a student is likely to express her or himself in ALL CAPS or with !!!!,
then he or she is likely to do it again, whereas other students never seem to use this type
of communication in threaded discussions.
Treating all uses of emotion equally raises other issues. For instance, students
may use the word “hope” but do not appear to be using it in an emotional sense.
Similarly, other students use the word “thanks” as a habitual salutation rather than as a
sign of appreciation.
Given this, it might be more useful for researchers to identify levels of each
indicator. For instance, researchers can identify instances of emotive text but then they
must identify whether it’s a strong, medium, or soft use. One way to address some of this
is to be able to interact with the participants as one codes the threaded discussions. In
other words, member checking might be an essential component when identifying social
presence behaviors because reading the text alone might not be enough. Or even better, a
researcher should be able to check both the original poster (about intent) as well as all
faculty and students about how they perceived the so-called social presence behavior
because analyzing online discussion behaviors without intent and how the
communication behaviors (i.e., the language used in the postings) are perceived is
limiting from both a design and a research perspective.

145. 130
Another problem—which I mentioned in Chapter 1—focuses on researchers’
tendency to treat all three categories and subsequent indicators of social presence equally.
As I mentioned in Chapter 2, some researchers tend to define social presence as not only
presenting oneself as “real” and “there” but also establishing a positive emotional
connection with others. In this case, it makes sense that while interactive and cohesive
types of communication are important and possibly necessary building blocks for
affective communication, affective communication is the best way to build an emotional
connection with others. In other words, simply ending a discussion posting with a
salutation is not near as powerful as disclosing personal information. Further research
though is needed to test this theory.
Problems with Measuring the Community of Inquiry
One final observation involves a conflict between the Community of Inquiry
Questionnaire—which was developed relatively recently by a team of CoI researchers
(Arbaugh et al., 2008; Swan et al., 2008)—and the indicators of social presence
developed by Rourke et al. (2001a). To illustrate my point, Table 5.3 lists the Community
of Inquiry Questionnaire questions for social presence next to Rourke et al.’s original
indicators.

146. 131
Table 5.3 Measuring Social Presence in a Community of Inquiry
Community of Inquiry Questionnaire Indicators of Social Presence
Affective expression
14. Getting to know other course participants
gave me a sense of belonging in the course.
15. I was able to form distinct impressions of
some course participants.
16. Online or web-based communication
is an excellent medium for social interaction.
--Paralanguage
--Emotion
--Humor
--Self Disclosure
Open communication
17. I felt comfortable conversing through
the online medium.
18. I felt comfortable participating in the
course discussions.
19. I felt comfortable interacting with
other course participants.
--Acknowledgement
--Agreement / Disagreement
--Invitation
--Expressing Appreciation
Group cohesion
20. I felt comfortable disagreeing with
other course participants while still
maintaining a sense of trust.
21. I felt that my point of view was
acknowledged by other course
participants.
22. Online discussions help me to develop a sense of
collaboration.
--Greetings & Salutations /
Phatics
--Vocatives
--Group Reference / Inclusivity
--Embracing the Group
While keeping in mind that the Community of Inquiry Questionnaire is meant to
measure student's attitudes and perceptions and the indicators of social presence are
meant to identify what students do and say, one would still expect to see more overlap
between the two instruments. But when I look at the questions for affective expression
and then the indicators for affective expression, I see very little overlap. First, the

147. Community of Inquiry Questionnaire has some strange questions that seem to focus more
on the medium than on what a participant does. For instance, question 16 states “online
or web-based communication is an excellent medium for social interaction.” Students are
132
asked the degree to which they agree with this statement. But is it not possible that
students might agree that CMC can be an excellent medium for social interaction but
disagree that CMC has been well used in a specific course? Then question 14 appears to
be inquiring about a student's sense of belonging but the social indicators do not address
that. In fact, group cohesion and specifically group reference seem like better indicators
of students’ feeling a sense of belonging to a group. The problems continue when one
looks at open communication and group cohesion. My point or rather my insight is that
researchers who use the Community of Inquiry Questionnaire to study social presence
appear to be studying different things than those who use indicators of social presence.
Limitations of the Study
As I mentioned earlier and in Chapter 1, every study suffers from some type of
limitation. Perhaps the first limitation of this study is the small sample size. While I
intentionally chose this small sample as a starting point for my line of research, I
recognize that multiple samples might have provided a nice point of comparison.
Threaded discussions are rich and full of data for researchers to mine. But I have
come to the conclusion that relying only on threaded discussions is limiting. A researcher
misses the things that might be said in emails, over the phone, or even in assignments
turned in to a drop-box. Recently, researchers (Archer, 2010; Shea et al., 2010; Shea et
al., 2009; Shea, Vickers, et al., 2009; Shea & Vickers, 2010) have argued about the need
to look at an entire course—rather than just threaded discussions or survey data—when

148. 133
studying the CoI. Like these researchers, I have found that another limitation of this
study is that it only focused on what was posted in the threaded discussions.
Last but not least, conducting content analysis without being able to check with
students about the meaning behind their postings as well as how other students interpret
their postings is also problematic and perhaps the biggest limitation of studies like this.
Concluding Thoughts and Implications
Despite the aforementioned limitations, the results of this study can be useful for
researchers and practitioners alike. From a research perspective, the study suggests that
social presence is much more complicated than previously conceptualized. While it is
helpful to investigate how students establish and maintain social presence in online
courses—which in this study was restricted to investigating postings in threaded
discussions—the list of indicators of social presence originally developed by Rourke et
al. (2001a) need to be revised. Further, multiple and mixed methods should be employed
whenever possible to investigate not only what students do and say but also how these
behaviors are perceived by others. Finally, and perhaps most importantly, researchers
need to spend more time focusing on how situational variables, such as the size of the
group, the instructional task, and the instructor’s role, in combination with personal
preferences influence how social presence is established and maintained.
Research, though, should inform practice. The results of this study, despite
limitations, have a number of pedagogical implications that instructional designers and
faculty alike can apply. For instance, the results of this study point to the importance of
the intentional use of different types of group activities and threaded discussions. Much
like in large lecture classrooms, students taking part in large threaded discussions might

149. feel that their voice is lost, that it is simply too hard to reach out and be heard, and that it
134
is too difficult to project one’s personality. Further, rather than having several small
weekly threaded discussions, it might make more sense—at least from a social presence
perspective—to include some longer project-based discussions with smaller groups that
take place over multiple weeks.
The results also argue for the importance of “positive” past relationships with
students. Developing a traditional cohort model where students complete their entire
program of study with the same group of students is one way to help leverage the power
of past relationships to build social presence. This research, though, simply suggests that
it might be advantageous in terms of social presence to have students take a few back-to-back
courses that build upon one another together, and involve the students in well
orchestrated group work. This research also seems to suggest that having the same
instructor teach more than one course (and possibly back to back) could be powerful in
terms of leveraging past relationships between instructors and students—this though
assumes the past relationships are positive.
However, when it is not possible to have students complete their program of study
in a cohort or take back-to-back courses, designers and faculty could focus on having
both getting-to-know-you activities upfront as well as reconnecting activities throughout
a given course (see Dunlap & Lowenthal, 2011, in press, for more on “getting to know
you” strategies that can be used in the beginning of a course as well as throughout a
course).
Finally, it is likely that a magic social presence formula does not exist. Each
student might have her or his own sensitivity to and proficiency at projecting her- or

150. himself as “real” and being “there” and specifically establishing connections with others.
135
Therefore, instructors and designers must try to find multiple and continual ways for
students and instructors to present themselves as “real” and “there” in threaded
discussions as well as other parts of online courses. For instance, some of the strategies I
have used and written about with colleagues to establish and maintain social presence
involve using digital stories (see Lowenthal & Dunlap, 2007, 2010), using social media
(see Dunlap & Lowenthal, 2009a, 2009b), using digital music (see Dunlap & Lowenthal,
2010b), giving feedback publicly (see Lowenthal & Thomas, 2010), and making one-on-one
phone calls to students (see Dunlap & Lowenthal, 2010a).
Given all of this, and in conclusion, designers and faculty should consider the
following elements the next time that they design or teach a course:
• Set up a variety of small-group discussion groups;
• Provide well structured small-group assignments that take time and
collaboration to complete;
• Balance instructor involvement;
• Establish incentives for students to take part in threaded discussions;
• Use a variety of instructional tasks and discussion prompts—some of which
ask for students to share personal and emotional details (when appropriate).

151. 136
APPENDIX A
Social Presence Measures
Table A1 Feelings about CMC
1. Stimulating-dull
2. Personal-impersonal
3. Sociable -unsociable
4. Sensitive-insensitive
5. Warm-cold
6. Colorful-colorless
7. Interesting-boring
8. Appealing-not appealing
9. Interactive-noninteractive
10. Active-passive
11. Reliable-unreliable
12. Humanizing-dehumanizing
13. Immediate-non-immediate
14. Easy-difficult
15. Efficient-inefficient
16. Unthreatening-threatening
17. Helpful-hindering
Note. From “Social Presence Theory and Implications for Interaction and Collaborative
Learning in Computer Conferences,” by C. N. Gunawardena, 1995, in International
Journal of Educational Telecommunications, 1(2/3), 147-166.

152. 137
Table A2 Social Presence Scale
1. Messages in GlobalEd were impersonal.
2. CMC is an excellent medium for social interaction.
3. I felt comfortable conversing through this text-based medium.
4. I felt comfortable introducing myself on GlobalEd.
5. The introduction enabled me to form a sense of online community.
6. I felt comfortable participating in GlobalEd discussions.
7. The moderators created a feeling of online community.
8. The moderators facilitated discussions in the GlobalEd conference.
9. Discussions using the medium of CMC tend to be more impersonal than face-to-
face discussion.
10. CMC discussions are more impersonal than audio conference discussions.
11. CMC discussions are more impersonal than video teleconference discussions.
12. I felt comfortable interacting with other participants in the conference.
13. I felt that my point of view was acknowledged by other participants in
GlobalEd.
14. I was able to form distinct individual impressions of some GlobalEd
participants even though we communicated only via a text-based medium.
Note. From “Social Presence as a Predictor of Satisfaction Within a Computer-mediated
Conferencing Environment,” by C. N. Gunawardena and F. J. Zittle, 1997, in The
American Journal of Distance Education, 11(3), 8-26.

153. 138
Table A3 Model and Template for Assessment of Social Presence
Category Indicators Definition of Indicators
Affective
Expression of
Responses
emotions
Conventional expressions of emotion, or
unconventional expressions of emotion,
includes repetitious punctuation,
conspicuous capitalization, emoticons
Use of Humor Teasing, cajoling, irony, understatements,
sarcasm
Self-Disclosure Presents details of life outside of class, or
expresses vulnerability
Interactive
Responses
Continuing a Thread Using reply feature of software, rather
than starting a new thread
Quoting from Other
Messages
Using software features to quote others
entire message or cutting and pasting
sections of others’ messages
Referring explicitly
to other messages
Direct references to contents of others’
posts
Asking questions Students ask questions of other students or
the moderator
Complimenting,
expressing
appreciation
Complimenting others or contents of
others’ messages
Expressing
agreement
Expressing agreement with others or
content of others’ messages
Cohesive
Responses
Vocatives Addressing or referring to participants by
name
Addresses or refers to
the group using
inclusive pronouns
Addresses the group as we, us, our, group
Phatics / Salutations Communication that serves a purely social
function; greetings, closures
Note. From “Assessing Social Presence in Asynchronous Text-based Computer
Conferencing,” by L. Rourke, D. R. Garrison, and W. Archer, 2001a, in Journal of
Distance Education, 14.

154. 139
Table A4 Additional Social Presence Variables
Dimensions
I. Social Context
II. Online
Communication
III. Interactivity
IV. Privacy
Familiarity with
recipients
Keyboarding and
accuracy skills
Timely Response Formats of CMC
Assertive /
acquiescent
Use of emoticons
and paralanguage
Communication
Styles
Access and
Location
Informal/formal
relationship
Characteristics of
real-time
discussion
Length of
Messages
Patterns of CMC
Trust relationships Characteristics of
discussion boards
Formal/Informal
Social
relationships (love
and information)
Language skills
(reading, writing)
Type of tasks
(planning,
creativity, social
tasks)
Psychological
attitude toward
technology
Size of Groups
Access and
location
Communication
strategies
User’s
characteristics
Note. From “The Relationship of Social Presence and Interaction in Online Classes,” by
C.-H. Tu and M. McIsaac, 2002, in The American Journal of Distance Education, 16(3),
131-150.

155. 140
Table A5 Community of Inquiry Survey Instrument
5 point Likert scale 1=strongly disagree, 2=disagree, 3=neutral, 4=agree, 5=strongly agree
Teaching Presence
Design & Organization
1. The instructor clearly communicated important course topics.
2. The instructor clearly communicated important course goals.
3. The instructor provided clear instructions on how to participate in course learning activities.
4. The instructor clearly communicated important due dates/time frames for learning activities.
Facilitation
5. The instructor was helpful in identifying areas of agreement and disagreement on course topics
that helped me to learn.
6. The instructor was helpful in guiding the class towards understanding course topics in a way
that helped me clarify my thinking.
7. The instructor helped to keep course participants engaged and participating in productive
dialogue.
8. The instructor helped keep the course participants on task in a way that helped me to learn.
9. The instructor encouraged course participants to explore new concepts in this course.
10. Instructor actions reinforced the development of a sense of community among course
participants.
Direct Instruction
11. The instructor helped to focus discussion on relevant issues in a way that helped me to learn.
12. The instructor provided feedback that helped me understand my strengths and weaknesses.
13. The instructor provided feedback in a timely fashion.
Social Presence
Affective expression
14. Getting to know other course participants gave me a sense of belonging in the course.
15. I was able to form distinct impressions of some course participants.
16. Online or web-based communication is an excellent medium for social interaction.
Open communication
17. I felt comfortable conversing through the online medium.
18. I felt comfortable participating in the course discussions.
19. I felt comfortable interacting with other course participants.
Group cohesion
20. I felt comfortable disagreeing with other course participants while still maintaining a sense of
trust.
21. I felt that my point of view was acknowledged by other course participants.
22. Online discussions help me to develop a sense of collaboration.

156. 141
Table A5 (con’t.)
Cognitive Presence
Triggering event
23. Problems posed increased my interest in course issues.
24. Course activities piqued my curiosity.
25. I felt motivated to explore content related questions.
Exploration
26. I utilized a variety of information sources to explore problems posed in this course.
27. Brainstorming and finding relevant information helped me resolve content related questions.
28. Online discussions were valuable in helping me appreciate different perspectives.
Integration
29. Combining new information helped me answer questions raised in course activities.
30. Learning activities helped me construct explanations/solutions.
31. Reflection on course content and discussions helped me understand fundamental
concepts in this class.
Resolution
32. I can describe ways to test and apply the knowledge created in this course.
33. I have developed solutions to course problems that can be applied in practice.
34. I can apply the knowledge created in this course to my work or other non-class related
activities.
Note. From “Validating a Measurement Tool of Presence in Online Communities of
Inquiry,” by K. Swan, P. Shea, J. Richardson, P. Ice, D. R. Garrison, M. Cleveland-Innes,
and J. B. Arbaugh, 2008, in E-Mentor, 2(24), 1-12.

157. 142
APPENDIX B
Word Count Results
Table B1 Word Count Results across All Forums
Rank Word Count Percentage (%)
1 I 4858 4.13
2 you 2186 1.86
3 have 1428 1.21
4 we 1367 1.16
5 my 1001 0.85
6 what 948 0.81
7 do 814 0.69
8 your 810 0.69
9 can 730 0.62
10 policy 600 0.51
11 me 595 0.51
12 all 592 0.50
13 about 574 0.49
14 bob 566 0.48
15 so 565 0.48
16 instructor 564 0.48
17 think 553 0.47
18 our 538 0.46
19 work 494 0.42
20 would 482 0.41
21 one 456 0.39
22 how 454 0.39
23 reading 428 0.36
24 week 421 0.36
25 from 414 0.35
26 some 414 0.35
27 more 407 0.35
28 just 392 0.33
29 know 390 0.33
30 need 383 0.33
31 get 381 0.32
32 group 348 0.30
33 doc 346 0.29
34 also 345 0.29
35 an 341 0.29
36 well 334 0.28
37 out 328 0.28
38 school 324 0.28
39 like 320 0.27
40 good 318 0.27
41 thanks 313 0.27
42 here 302 0.26
43 which 288 0.24
44 other 281 0.24
45 should 275 0.23
46 data 264 0.22
47 i’m 264 0.22
48 paper 261 0.22
49 could 254 0.22
50 research 251 0.21

158. 143
Table B2 Word Count Results across Project Groups
Rank Word Count Percentage (%)
1 I 1674 4.08
2 you 729 1.78
3 we 678 1.65
4 have 497 1.21
5 what 387 0.94
6 do 267 0.65
7 can 261 0.64
8 your 258 0.63
9 may 255 0.62
10 our 251 0.61
11 all 241 0.59
12 think 221 0.54
13 my 215 0.52
14 so 214 0.52
15 data 202 0.49
16 policy 193 0.47
17 would 184 0.45
18 some 181 0.44
19 need 176 0.43
20 me 173 0.42
21 about 171 0.42
22 from 169 0.41
23 paper 169 0.41
24 work 166 0.40
25 how 161 0.39
26 draft 155 0.38
27 just 150 0.37
28 know 150 0.37
29 also 144 0.35
30 bob 142 0.35
31 instructor 142 0.35
32 more 139 0.34
33 out 136 0.33
34 should 133 0.32
35 thanks 132 0.32
36 get 130 0.32
37 schools 127 0.31
38 mary 125 0.30
39 here 122 0.30
40 good 120 0.29
41 doc 118 0.29
42 like 114 0.28
43 could 112 0.27
44 group 112 0.27
45 make 111 0.27
46 project 111 0.27
47 week 111 0.27
48 school 109 0.27
49 programs 108 0.26
50 well 107 0.26

159. 144
Table B3 Word Count Results across Pairs
Rank Word Count Percentage (%)
1 I 960 4.87
2 you 438 2.22
3 my 339 1.72
4 have 291 1.48
5 we 209 1.06
6 your 189 0.96
7 me 148 0.75
8 what 145 0.74
9 do 130 0.66
10 work 123 0.62
11 about 119 0.60
12 goals 116 0.59
13 can 110 0.56
14 our 102 0.52
15 how 100 0.51
16 school 97 0.49
17 time 88 0.45
18 teachers 85 0.43
19 goal 84 0.43
20 some 82 0.42
21 would 82 0.42
22 think 79 0.40
23 bob 76 0.39
24 instructor 76 0.39
25 need 75 0.38
26 know 73 0.37
27 so 72 0.37
28 an 71 0.36
29 week 66 0.34
30 well 65 0.33
31 all 63 0.32
32 out 62 0.31
33 been 61 0.31
34 like 61 0.31
35 more 60 0.30
36 one 60 0.30
37 also 59 0.30
38 from 58 0.29
39 just 57 0.29
40 get 55 0.28
41 see 55 0.28
42 good 54 0.27
43 when 54 0.27
44 where 54 0.27
45 i’m 53 0.27
46 each 52 0.26
47 help 52 0.26
48 students 52 0.26
49 other 51 0.26
50 plan 50 0.25

160. 145
Table B4 Word Count Results across Reading Groups
Rank Word Count Percentage (%)
1 I 1784 3.79
2 you 802 1.70
3 have 532 1.13
4 we 416 0.88
5 what 358 0.76
6 do 348 0.74
7 policy 344 0.73
8 my 328 0.70
9 can 297 0.63
10 reading 293 0.62
11 your 283 0.60
12 one 255 0.54
13 about 242 0.51
14 all 231 0.49
15 think 227 0.48
16 me 226 0.48
17 so 225 0.48
18 instructor 222 0.47
19 bob 221 0.47
20 week 202 0.43
21 summary 196 0.42
22 doc 190 0.40
23 would 185 0.39
24 more 175 0.37
25 how 171 0.36
26 our 165 0.35
27 get 164 0.35
28 from 162 0.34
29 questions 161 0.34
30 work 156 0.33
31 just 155 0.33
32 log 151 0.32
33 group 141 0.30
34 which 141 0.30
35 an 139 0.29
36 know 138 0.29
37 research 138 0.29
38 some 138 0.29
39 well 131 0.28
40 democratic 130 0.28
41 good 128 0.27
42 Lasswell 128 0.27
43 also 126 0.27
44 chapter 122 0.26
45 here 122 0.26
46 has 117 0.25
47 like 115 0.24
48 thanks 114 0.24
49 heck 113 0.24
50 any 112 0.24

161. 146
APPENDIX C
Constant Comparison Analysis Results
Table C.1 Reading Group E Codes
Codes Generated
Acknowledging lack of
knowledge
Greeting Reflection
Addressing question Happiness Reflection about course
material
Advice Heading Relating
Agreement Hope Relating to Others
Answer Hoping for help Research discussion
Answering questions Humor Resource
Anticipation Justification of example Resource recommendation
Apology Likes course reading Response
Appreciation Note Reveal life outside of class
Assignment discussion Opinion Reveal life outside of class as
relates to class
Belief Paralanguage Reveal problems
Bias Personal course interests Reveal struggling
Clarification Personal example Reveals lack of knowledge
Commitment to more
Personal interest in course stuff Salutation
discussion
Complimenting texts Personal interest in reading Shares thinking about thinking
Contextualizing Point Personal Life Details Shares thoughts about reading
Critique Personal story Sharing thoughts
Critique of writing Personal study details Sharing values
Discussing Reading Personalization of Material Showing relevance
Discussing policy Philosophical Discussion Story example
Doubt Plan Thinking about reading
Empathy Plea Thinking about the course
Enjoyment Policy answer Thinking about thinking
Example Positive feedback Thought
Example of criticism Positive thinking Thoughts about policy
Excitement Question Thoughts about reading
Explaining struggles Quotation Vocative
Explanation Reading discussion Wonder
General Policy Discussion Recommendation Worth mentioning
Grade details Recommending other sources

162. 147
Table C.2 Reading Group E Groups
Grouping of Codes
Course logistics & facilitation
Addressing question
Answering questions
Critique
Critique of writing
Grade details
Heading
Question
Advice
Recommendation
Recommending other sources
Policy Related Class Discussions
Reflection
Reflection about course material
Thinking about reading
Thinking about the course
Thinking about thinking
General Policy Discussion
Discussing reading
Discussing policy
Thoughts about policy
Thoughts about reading
Policy answer
Shares thinking about thinking
Shares thoughts about reading
Sharing thoughts
Research discussion
Showing relevance
Belief
Bias
Thought
Wonder
Worth mentioning
Response
Resource
Resource recommendation
Personal interest in course stuff
Personal interest in reading
Plan
Personalization of Material
Philosophical Discussion
Complimenting texts
Contextualizing Point
Justification of example
Likes course reading
Note
Opinion
Assignment discussion
Quotation
Reading discussion
Example
Example of criticism
Clarification
Commitment to more discussion
Personal course interests
Answer
Explanation
Emotion
Anticipation
Paralanguage
Apology
Doubt
Empathy
Enjoyment
Excitement
Happiness
Hope
Hoping for help
Plea
Humor
Greetings and Salutations
Greeting
Salutation
Vocative
Sharing Life Details
Personal example
Personal story
Story example
Reveal life outside of class
Reveal life outside of class as relates to class
Personal Life Details
Personal study details
Gracious / Gratitude
Appreciation
Positive feedback
Positive thinking
Self Disclosing Personal Matters
Explaining struggles
Acknowledging lack of knowledge
Reveal problems
Reveal struggling
Reveals lack of knowledge
Sharing values
Playing Nice with Others
Agreement
Relating
Relating to Others

163. 148
Table C.3 Pair9 Codes
Codes Generated
Acknowledgement Positive feedback
Agreement Positive self assessment
Answer to question Question
Anticipation Question about meeting
Appreciation Reassurance
Assignment details Recommendation
Assignment discussion Reference education literature
Best wishes Reflection
Brainstorming Relating
Commitment Relating to others
Concern Request
Confidence in peers Reveal uncertainty
Course planning Revealing concerns
Doubt Revealing life in other courses
Emotion Revealing struggles
Enjoyment Revealing thinking
Explanation Revealing unawareness
Explanation for struggles Salutation
Feeling overwhelmed Self assessment
Greeting Self disclosure
Happiness Sharing course plans
Heading Sharing plans
Hope Sharing successes
Inquiring about life outside of course Thanks
Introspection Thinking about policy
invitation Thinking out loud
Likes idea Thoughts on assignments
Paralanguage Thoughts on instructor
Personal sharing Thoughts on leadership
Plan to collaborate Understanding task
Plans Vocative
Plans to meet

164. 149
Table C.4 Pair9 Groups
Course logistics & facilitation Collaboration
Recommendation
Question
Request
invitation
Understanding task
Assignment details
Assignment discussion
Heading
Question about meeting
Course planning
Plan to collaborate
Plans
Plans to meet
Sharing course plans
Sharing plans
Emotion Sharing Life Details
Concern
Anticipation
Happiness
Doubt
Emotion
Enjoyment
Feeling overwhelmed
Hope
Paralanguage
Introspection
Personal sharing
Reflection
Self assessment
Revealing life in other courses
Positive self assessment
Sharing successes
Inquiring about life outside of course
Playing Nice with Others Policy Related Class Discussions
Relating
Reference education literature
Relating to others
Brainstorming
Acknowledgement
Thinking about policy
Agreement
Thinking out loud
Answer to question
Thoughts on assignments
Likes idea
Thoughts on instructor
Confidence in peers
Thoughts on leadership
Commitment
Revealing thinking
Reassurance
Explanation
Self Disclosing Personal Matters Gracious / Gratitude
Revealing unawareness
Revealing struggles
Self disclosure
Reveal uncertainty
Revealing concerns
Explanation for struggles
Positive feedback
Best wishes
Appreciation
Thanks
Greetings and Salutations
Greeting
Vocative
Salutation

165. 150
REFERENCES
Akyol , Z., & Garrison, D.R. (2009). Community of inquiry in adult online learning:
Collaborative-constructivist approaches. In T. T. Kidd (Ed.), Adult learning in the
digital age: Perspectives on online technologies and outcomes (pp. 55-66).
Hershey, PA: IGI Global.
Akyol, Z., Vaughan, N., & Garrison, D.R. (2011). The impact of course duration on
the development of a community of inquiry. Interactive Learning Environments,
19(3), 231-246.
Ali, R., & Leeds, E. M. (2010). The impact of face-to-face orientation on online
retention: A pilot study. Online Journal of Distance Learning Administration,
13(4). Retrieved from http://www.westga.edu/~distance/ojdla/winter124
/ali124.html
Allen, I. E., & Seaman, J. (2006). Making the grade: Online education in the United
States, 2006. Needham, MA: Sloan-C.
Allen, I. E., & Seaman, J. (2010). Learning on demand: Online education in the United
States, 2009. Needham, MA: Sloan-C.
An, H., Shin, S., & Lim, K. (2009). The effects of different instructor facilitation
approaches on students’ interactions during asynchronous online discussions.
Computers & Education, 53(3), 749-760.
Anderson, T. (2003a). Getting the mix right again: An updated and theoretical
rationale for interaction. The International Review of Research in Open and
Distance Learning, 4(2). Retrieved from http://www.irrodl.org/index.php/
irrodl/article/view/149/230

166. Anderson, T. (2003b). Modes of interaction in distance education: Recent developments
and research questions. In M. Moore (Ed.), Handbook of distance education (pp.
151
129-144). Mahwah, NJ: Lawrence Erlbaum.
Anderson, T., & Garrison, D. R. (1998). Learning in a networked world: New roles and
responsibilities. In C. Gibson (Ed.), Distance learners in higher education. (pp.
97-112). Madison, WI: Atwood.
Anderson, T., & Kuskis, A. (2007). Modes of interaction. In M. G. Moore (Ed.),
Handbook of distance education (pp. 295-309). Mahwah, NJ: Lawrence Erlbaum.
Aragon, S. R. (2003). Creating social presence in online environments. New Directions
for Adult and Continuing Education, 100, 57-68.
Arbaugh, J.B. (2007). An empirical verification of the community of inquiry
framework. Journal of Asynchronous Learning Network, 11(1), 73-85.
Arbaugh, B., Bangert, A., & Cleveland-Innes, M. (2010). Subject matter effects and the
community of inquiry framework. The Internet and Higher Education, 13(1-2),
37-44.
Arbaugh, J. B., & Benbunan-fich, R. (2006). An investigation of epistemological and
social dimensions of teaching in online learning environments. Academy of
Management Learning & Education, 5(4), 435-447.
Arbaugh, J. B., Cleveland-Innes, M., Diaz, S. R., Garrison, G. R., Philip, I.,
Richardson, J. C., Shea, P., & Swan, K. P. (2008). The community of inquiry
framework: Development, validation, and directions for further research. Paper
presented at the annual meeting of the American Education Research Association,
New York, NY.

167. 152
Argyle, M. (1969). Social interaction. New York: Atherton Press.
Argyle, M., & Cook, M. (1976). Gaze and mutual gaze. London: Cambridge University.
Argyle, M., & Dean, J. (1965). Eye contact, distance and affiliation. Sociometry, 28, 289-
304.
Benbunan-Fich, R., Hiltz, S. R., & Harasim, L. (2005). The online interaction learning
model: An integrated theoretical framework for learning networks. In S. R. Hiltz
& R. Goldman (Eds.), Learning together online: Research on asynchronous
learning networks (pp. 19-37). Mahwah, NJ: Lawrence Erlbaum.
Berelson, B. (1952). Content analysis in communication research. Glencoe, IL: Free
Press.
Bernard, R. M., Abrami, P. C., Lou, Y., Borokhovski, E., Wade, A., Wozney, L., et
al. (2004). How does distance education compare with classroom instruction? A
meta-analysis of the empirical literature. Review of Educational Research, 74(3),
379-439.
Biocca, F. (1997). The cyborg's dilemma: Progressive embodiment in virtual
environments. Journal of Computer-Mediated Communication, 3(2). Retrieved
from http://www. ascusc.org/jcmc/vol3/issue2/biocca2.html
Biocca, F., & Harms, C. (2002). What is social presence? In F. Gouveia & F. Biocca,
Presence 2002 Proceedings. Porto, Portugal: University of Fernando Pessoa
Press.
Biocca, F., Harms, C., & Burgoon, J. K. (2003). Toward a more robust theory and
measure of social presence: Review and suggested criteria. Presence:
Teleoperators & Virtual Environments, 12(5), 456-480.

168. Brown, J. S. (2002). Growing up digital: How the web changes work, education, and the
153
ways people learn. USDLA Journal, 16(2). Retrieved
fromhttp://www.usdla.org/html/journal/FEB02_Issue/article01.html
Carley, K. (1993). Coding choices for textual analysis: A comparison of content analysis
and map analysis. Sociological Methodology, 23, 75-126.
Caspi, A., & Blau, I. (2008). Social presence in online discussion groups: Testing
three conceptions and their relations to perceived learning. Social Psychology
of Education, 11(3), 323-346.
Chickering, A.W., & Gamson, Z.F. (1987). Seven principles for good practice in
undergraduate education. AAHE Bulletin, 39(7), 3-7.
Christie, B. (1974). Perceived usefulness of person-person telecommunications media
as a function of the intended application. European Journal of Social Psychology,
4(3), 366-368.
Christie, B., & Holloway, S. (1975). Factors affecting the use of telecommunications
by management. Journal of Occupational Psychology, 48, 3-9.
Christie, B., & Kingan, S. (1977). Electronic alternatives to the business meeting:
Managers’ choices. Journal of Occupational Psychology, 50, 265-273.
Collins, K. M. T., Onwuegbuzie, A. J., & Sutton, I. L. (2006). A model incorporating
the rationale and purpose for conducting mixed methods research in special
education and beyond. Learning Disabilities: A Contemporary Journal, 4, 67-
100.
Connolly, T., Jessup, L. M., & Valacich, J, S, (1990). Effects of anonymity and

169. 154
evaluative tone on idea generation in computer-mediated groups. Management
Science, 36, 97-120.
Creswell, J. W. (1998). Qualitative inquiry and research design: Choosing among
five traditions. Thousand Oak, CA: Sage.
Creswell, J. W. (2008). Educational research: Planning, conducing, and evaluating
quantitative and qualitative research (3rd ed.). Upper Saddle River, NJ: Pearson.
Creswell, J. W., & Plano Clark, V. L. (2007). Designing and conducting mixed methods
research. Thousand Oaks, CA: Sage.
Daft, R. L., & Lengel, R. H. (1984). Information richness: A new approach to
managerial behavior and organizational design. In L. L. Cummings & B. M.
Staw (Eds.), Research in organizational behavior (pp. 191-233). Homewood, IL:
JAI Press.
Daft, R. L., & Lengel, R. H. (1986). Organizational information requirements, media
richness and structural design. Management Science, 32(5), 554-571.
Daft, R. L., Lengel, R. H., & Trevino, L. K. (1987). Message equivocality, media
selection, and manager performance: Implications for information systems. MIS
Quarterly, 11(3), 355-366.
Danchak, M. M., Walther, J. B., & Swan, K. P. (2001, November). Presence in mediated
instruction: Bandwidth, behavior, and expectancy violations. Paper presented at
the annual meeting of Asynchronous Learning Networks, Orlando, FL.
De Wever, B., Schellens, T., Valcke, M., & Van Keer, H. (2006). Content analysis
schemes to analyze transcripts of online asynchronous discussion groups: A
review. Computers & Education, 46, 6-28.

170. 155
Delfino M., & Manca S. (2007). The expression of social presence through the use of
figurative language in a web-based learning environment. Computers in Human
Behaviour, 23(5), 2190-2211
Dellinger, A. B., & Leech, N. L. (2007). Toward a unified validation framework in mixed
methods research. Journal of Mixed Methods Research, 1(4), 309-332.
Dennen, V. P. (2005). From message posting to learning dialogues: Factors affecting
learner participation in asynchronous discussion. Distance Education, 26(1), 127-
148.
Dennen, V. P. (2008). Pedagogical lurking: Student engagement in non-posting
discussion behavior. Computers in Human Behavior, 24(4), 1624-1633.
Dunlap, J. C., & Lowenthal, P. R. (2009a). Horton hears a tweet. EDUCAUSE
Quarterly, 32(4).
Dunlap, J. C., & Lowenthal, P. R. (2009b). Tweeting the night away: Using Twitter to
enhance social presence. Journal of Information Systems Education, 20(2), 129-
136.
Dunlap, J. C., & Lowenthal, P. R. (2010a). Defeating the Kobayashi Maru: Supporting
student retention by balancing the needs of the many and the one. EDUCAUSE
Quarterly, 33(4). Retrieved from http://www.educause.edu/
EDUCAUSE+Quarterly/EDUCAUSEQuarterlyMagazineVolum/DefeatingtheKo
bayashiMaruSuppo/219103
Dunlap, J. C., & Lowenthal, P. R. (2010b). Hot for teacher: Using digital music to
enhance student’s experience in online courses. TechTrends, 54(4), 58-73. doi:
10.1007/s11528-010-0421-4

171. 156
Dunlap, J. C., & Lowenthal, P. R. (2011). Who are you? Alternative online meet-and-
greet tactics. In P. Shank (Ed.), The online learning idea book: Proven ways
to enhance technology-based and blended learning (vol. 2; pp. 149-152). San
Francisco: Pfeiffer.
Dunlap, J. C., & Lowenthal, P. R. (in press). Getting to know you: The first week of
class and beyond. In P. R. Lowenthal, D. Thomas, B. Yuhnke, A. Thai, M.
Edwards, & C. Gasell (Eds.), The CU Online Handbook, 2011. Raleigh, NC: Lulu
Enterprises.
Dunlap, J. C., Sobel, D., & Sands, D. I. (2007). Supporting students’ cognitive processing
in online courses: Designing for deep and meaningful student-to-content
interactions. TechTrends, 51(4), 20-31.
Eisenhart, M., & Howe, K. (1992). Validity in educational research. In M. LeCompte,
W. Millroy, & J. Preissle (Eds.), The handbook of qualitative research in
education (pp. 642-680). San Diego, CA: Academic Press.
Fryshman, B. (2008). Do we assess learning? Pull up a chair… Inside Higher Ed.
Retrieved from http://www.insidehighered.com/views/2008/08/07/fryshman
Garrison, D. R., & Anderson, T. (2003). E-learning in the 21st century: A framework for
research and practice. New York: RoutledgeFalmer.
Garrison, D. R. & Arbaugh, J.B. (2007). Researching the community of Inquiry
Framework: Review, Issues, and Future Directions. The Internet and Higher
Education, 10(3), 157-172.

172. 157
Garrison, D. R., Anderson, T., & Archer, W. (2000). Critical inquiry in a text-based
environment: Computer conferencing in higher education. The Internet and
Higher Education, 2(2-3), 87-105.
Garrison, D. R., Anderson, T., & Archer, W. (2001). Critical thinking, cognitive
presence, and computer conferencing in distance education. American Journal of
Distance Education, 15(1), 7-23.
Garrison, D. R., Cleveland-Innes, M., Koole, M., & Kappelman, J. (2006). Revisiting
methodological issues in transcript analysis: Negotiated coding and reliability.
Internet and Higher Education, 9(1), 1-8.
Garrison, D. R., Cleveland-Innes, M., & Fung, T. S. (2010). Exploring causal
relationships among teaching, cognitive, and social presence: Student perceptions
of the community of inquiry framework. The Internet and Higher Education,
13(1-2), 31-36.
Gilbert, P. K., & Dabbagh, N. (2005). How to structure online discussions for
meaningful discourse: A case study. British Journal of Educational Technology,
36(1), 5-18.
Glaser, B. G., & Strauss, A. L. (1967). The discovery of grounded theory: Strategies for
qualitative research. New Brunswick, NJ: AldineTransaction.
Goldman, R., Crosby, M., Swan, K., & Shea, P. (2005). Qualitative and quisitive research
methods for describing online learning. In S. R. Hiltz & R. Goldman, Learning
together online (pp. 103-120). Mahwah, NJ: Lawrence Erlbaum.
Goodwin, L. D. (2001). Interrater agreement and reliability. Measurement in Physical
Education and Exercise Science, 5(1), 13-34.

173. Goodwin, L. D., & Leech, N. L. (2003). The meaning of validity in the new standards for
158
educational and psychological testing: Implications for measurement courses.
Measurement and Evaluation in Counseling and Development, 36(3), 181-191.
Gorsky, P., Caspi, A. Antonovsky, A., Blau, I., & Mansur, A. (2010). The relationship
between academic discipline and dialogic behavior in open university course
forums. The International Review of Research in Open and Distance Learning,
11(2). Retrieved from http://www.irrodl.org/index.php/
irrodl/article/view/820/1546
Gunawardena, C. N. (1995). Social presence theory and implications for interaction and
collaborative learning in computer conferences. International Journal of
Educational Telecommunications, 1(2/3), 147-166.
Gunawardena, C. N., & Zittle, F. J. (1997). Social presence as a predictor of satisfaction
within a computer-mediated conferencing environment. The American Journal of
Distance Education, 11(3), 8-26.
Gunawardena, C. N., Lowe, C. A., & Anderson, T. (1997). Analysis of a global online
debate and the development of an interaction analysis model for examining social
construction of knowledge in computer conferencing. Journal of Educational
Computing Research, 17(4), 397-431.
Henninger, M., & Viswanathan, V. (2004). Social presence in online tutoring: What we
know and what we should know. In P. Gerjets, P. A. Kirschner, J. Elen, & R.
Joiner (Eds.), Proceedings of the first joint meeting of the EARLI SIGs
Instructional Design and Learning and Instruction with Computers (CD-ROM)
(pp. 365-376). Tuebingen: Knowledge Media Research Center.

174. 159
Henri, F. (1992). Computer conferencing and content analysis. In A. R. Kaye (Ed.),
Collaborative learning through computer conferencing. The Najadan Papers (pp.
117-136). London: Springer-Verlag.
Herring, S. C. (2004). Content analysis for new media: Rethinking the paradigm. In New
research for new media: Innovative research methodologies symposium working
papers and readings (pp. 47-66). Minneapolis, MN: University of Minnesota
School of Journalism and Mass Communication.
Herring, S. C. (2007). A faceted classification scheme for computer-mediated
discourse. Language@Internet, 4(1). Retrieved from http://www.
languageatinternet.de/articles/2007/761/Faceted_Classification_Scheme_for_CM
D.pdf
Hiemstra, G. (1982). Teleconferencing, concern for face, and organizational culture.
In M. Burgoon (Ed.), Communication yearbook 6 (pp. 874-904). Newbury Park,
CA: Sage.
Hiltz, S. R., & Arbaugh, J. B. (2003). Improving quantitative research methods in studies
of asynchronous learning networks (ALN). In J. R. Bourne & J. C. Moore (Eds.),
Elements of quality online education: Practice and direction (pp. 59-72).
Needham, MA: Sloan-C.
Hiltz, S. R., & Turoff, M. (1993). The network nation. Cambridge, MA: MIT Press.
Holsti, O. R. (1969). Content analysis for the social sciences and humanities. Reading:
MA, Addison-Wesley.

175. Hostetter, C., & Busch, M. (2006). Measuring up online: The relationship between social
160
presence and student learning satisfaction. Journal of Scholarship of Teaching
and Learning, 6(2), 1-12.
Hughes, M., Ventura, S., & Dando, M. (2007). Assessing social presence in online
discussion groups: A replication study. Innovations in Education and Teaching
International, 44(1), 17-29.
Ice, P., Gibson, A. M., Boston, W. & Becher, D. (2011). An exploration of
differences between community of indicators in low and high disenrollment
online courses. Journal of Asynchronous Learning Networks, 15(2), 44-70.
Johansen, R. (1977). Social evaluations of teleconferencing. Telecommunications
Policy, 1(5), 395-419.
Joo, Y. J., Lim, K. Y. & Kim, E. K. (2011). Online university students’ satisfaction
and persistence: Examining perceived level of presence, usefulness and ease of
use as predictors in a structural model. Computers & Education, 57, 1654–1664.
Kear, K. (2010). Social presence in online learning communities. In L. Dirckinck-
Holmfeld, V. Hodgson , C. Jones, M. de Laat, D. McConnell, & T. Ryberg
(Eds.), Proceedings of the 7th International Conference on Networked Learning
(pp. 541-548). Lancaster: University of Lancaster. Retrieved from
http://www.lancs.ac.uk/fss/organisations/netlc/past/nlc2010/abstracts/PDFs/Kear.
pdf
Kemp, N. J., & Rutter, D. R. (1986). Social interaction in blind people: An
experimental analysis. Human Relations, 39(3), 195-210.
Kiesler, S. (1986). The hidden messages in computer networks. Harvard Business

176. 161
Review, 64(3), 46-54, 58-60.
Kiesler, S., Siegel, J., McGuire, T. W. (1984). Social psychological aspects of
computer-mediated communication. American Psychologist, 39(10), 1123-
1134.
Kramer, A. D. I., Oh, L. M., & Fussell, S. R. (2006). Using linguistic features to
measure presence in computer-mediated communication. In Proceedings of the
SIGCHI conference on Human Factors in Computing Systems (pp. 913-916). New
York: ACM Press.
Krathwohl, D. R. (2004). Methods of educational and social science research: An
integrated approach (2nd ed.). Long Grove, IL: Waveland Press.
Kreijns, K., Kirschner, P. A., & Jochems, W. (2003). Identifying the pitfalls of social
interaction in computer-supported collaborative learning environments: A review
of the research. Computers in Human Behavior, 19(3), 335-353.
Lane, C-A (2011, April). Social presence impacting cognitive learning of adults in
distance education. A thesis submitted to the faculty at Athabasca.
https://dt.athabascau.ca/jspui/bitstream/10791/10/1/MasterThesisLane-v36-
Apr282011.pdf
Leech, N. L., & Onwuegbuzie, A. J. (2006). A typology of mixed research designs.
Quality and Quantity. Retrieved January 22, 2008. doi: 10.1007/s11135-007-
9105-3
Leech, N. L., & Onwuegbuzie, A. J. (2007). An array of qualitative data analysis tools: A
call for data analysis triangulation. School Psychology Quarterly, 22(4), 557-584.

177. Levy, Y. (2007). Comparing dropouts and persistence in e-learning courses. Computers
162
& Education, 48(2), 185-204.
Lin, G.-Y. (2004, October). Social presence questionnaire of online collaborative
learning: Development and validity. Paper presented at the annual meeting of the
Association for Educational Communications and Technology, Chicago, IL.
Lombard, M., & Ditton, T. (1997). At the heart of it all: The concept of presence. Journal
of Computer-Mediated Communication, 3(2). Retrieved from
http://jcmc.indiana.edu/vol3/issue2/lombard.html
Lomicka, L. & Lord, G. (2007). Social presence in virtual communities of foreign
language (FL) teachers. System, 35, 208-228.
Lowenthal, P. R. (2010). The evolution and influence of social presence theory on online
learning. In S. Dasgupta (Ed.), Social computing: Concepts, methodologies, tools,
and applications (pp. 113-128). Hershey, PA: IGI Global.
Lowenthal, P. R. (2009). Social presence. In P. Rogers, G. Berg, J. Boettcher, C.
Howard, L. Justice, & K. Schenk (Eds.), Encyclopedia of distance and online
learning (2nd ed., pp. 1900-1906). Hershey, PA: IGI Global.
Lowenthal, P., & Dunlap, J. (2007). Digital stories. In P. Shank (Ed.), The online
learning idea book: 95 proven ways to enhance technology-based and blended
learning (pp. 110-111). San Francisco: Pfeiffer.
Lowenthal, P. R., & Dunlap, J. (2010). From pixel on a screen to real person in your
students’ lives: Establishing social presence using digital storytelling. The Internet
and Higher Education, 13(1-2), 70-72. doi:10.1016/j.iheduc.2009.10.004
Lowenthal, P. R., & Dunlap, J. C. (2011, April). Investigating students’ perceptions

178. of various instructional strategies to establish social presence. Paper presented at
163
the annual meeting of the American Educational Research Association, New
Orleans, LA.
Lowenthal, P. R., & Leech, N. (2009). Mixed research and online learning: Strategies
for improvement. In T. T. Kidd (Ed.), Online education and adult learning: New
frontiers for teaching practices (pp. 202-211). Hershey, PA: IGI Global.
Lowenthal, D. A., & Lowenthal, P. R. (2010, April). A mixed methods examination
of instructor social presence in accelerated online courses. Paper presented at the
annual meeting of the American Education Research Association, Denver, CO.
Lowenthal, P. R., Lowenthal, D. A., & White, J. W. (2009). The changing nature of
online communities of inquiry: An analysis of how discourse and time shapes
students' perceptions of presence. In M. Simonson (Ed.), 32nd Annual
proceedings: Selected research and development papers presented at the annual
convention of the Association for Educational Communications and Technology.
Washington DC: Association for Educational Communications and Technology.
Lowenthal, P. R., & Thomas, D. (2010). Death to the Digital Dropbox: Rethinking
student privacy and public performance. EDUCAUSE Quarterly, 33(3). Retrieved
from http://www.educause.edu/EDUCAUSE+Quarterly/
EDUCAUSEQuarterlyMagazineVolum/DeathtotheDigitalDropboxRethin/213672
Lowenthal, P. R., Wilson, B., & Parrish, P. (2009). Context matters: A description
and typology of the online learning landscape. In M. Simonson (Ed.), 32nd
Annual proceedings: Selected research and development papers presented at
the annual convention of the Association for Educational Communications

179. 164
and Technology. Washington D. C.: Association for Educational
Communications and Technology.
Ludwig-Hardman, S., & Dunlap, J. C. (2003). Learner support services for online
students: Scaffolding for success. The International Review of Research in Open
and Distance Learning, 4(1). Retrieved from http://www.irrodl.org
/index.php/irrodl/article/view/131/211
McIsaac, M. S., Blocher, J. M., Mahes, V., & Vrasidas, C. (1999). Student and teacher
perceptions of interaction in online computer-mediated communication.
Educational Media International, 36(2), 121-131.
Mehrabian, A. (1972). Nonverbal communication. New Brunswick, NJ:
AldineTransaction.
Miles, M., & Huberman, A. M. (1994). Qualitative data analysis: An expanded
sourcebook (2nd ed.). Thousand Oaks, CA: Sage.
Moore, M. G. (1989). Three types of interaction. The American Journal of Distance
Education, 3(2), 1-6.
Moore, M. G., & Kearsley, G. (2005). Distance education: A systems view (2nd ed.).
New York: Wadsworth.
Na Ubon, A. & Kimble, C. (2003. Supporting the creation of social presence in online
learning communities using asynchronous text-based CMC. In The 3rd
International Conference on Technology in Teaching and Learning in Higher
Education (pp. 295-300). Heidelberg, Germany.

180. National Center for Education Statistics. (2008). Distance education at degree-granting
165
postsecondary institutions: 2006-07. Washington, DC: U.S. Department of
Education. Retrieved from http://nces.ed.gov/pubs2009/
2009044.pdf
Nippard, E. & Murphy, E. (2007). Social Presence in the Web-based Synchronous
Secondary Classroom. Canadian Journal of Learning and Technology, 33(1).
Onwuegbuzie, A. J., & Collins, K. M. T. (2007). A typology of mixed methods sampling
designs in social science research. The Qualitative Report, 12(2), 281-316.
Onwuegbuzie, A. J., & Leech, N. L. (2004). Enhancing the interpretation of “significant”
findings: The role of mixed methods research. The Qualitative Report, 9(4), 770-
792.
Onwuegbuzie, A. J., & Leech, N. L. (2005b). Taking the “Q” out of research: Teaching
research methodology courses without the divide between quantitative and
qualitative paradigms. Quality & Quantity, 39, 267-296.
Onwuegbuzie, A. J., & Leech, N. L. (2007). Sampling designers in qualitative research:
Making the sampling process more public. The Qualitative Report, 12(2), 238-
254.
Pena-Shaff, J. B., & Nicholls, C. (2004). Analyzing student interactions and meaning
construction in computer bulletin board discussions. Computers & Education, 42,
243-265.
Picciano, A. (2002). Beyond student perceptions: Issues of interaction, presence, and
performance in an online course. Journal of Asynchronous Learning Networks,
6(1), 21-40.

181. 166
Prensky, M. (2001, September/October). Digital natives, digital immigrants. On the
Horizon, 9(5), 1-6.
Pye, R., & Williams, E. (1978). Teleconferencing: Is video valuable or is audio
adequate? Telecommunications Policy, 1(3), 230-241.
Rettie, R. (2003). Connectedness, awareness, and social presence. Paper presented at the
6th International Presence Workshop, Aalborg, Denmark.
Richardson, J. C., & Swan, K. (2003). Examining social presence in online courses in
relation to students' perceived learning and satisfaction. Journal of Asynchronous
Learning Networks, 7(1), 68-88.
Rourke, L., & Anderson, T. (2002a). Exploring social communication in computer
conferencing. Journal of Interactive Learning Research, 13(3), 259-275.
Rourke, L., & Anderson, T. (2002b). Using peer teams to lead online discussions. Journal
of Interactive Media in Education, 1, 1-21.
Rourke, L., & Anderson, T. (2004). Validity in qualitative content analysis. Educational
Technology Research and Development, 52(1), 5-18.
Rourke, L., Anderson, T., Garrison, D. R., & Archer, W. (2001a). Assessing social
presence in asynchronous text-based computer conferencing. Journal of Distance
Education, 14. Retrieved from http://cade.athabascau.ca/vol14.2/
rourke_et_al.html
Rourke, L., Anderson, T., Garrison, D. R., & Archer, W. (2001b). Methodological issues
in the content analysis of computer mediated transcripts. International Journal of
Artificial Intelligence in Education, 12, 8-22.
Rourke, L., & Kanuka, H. (2009). Learning in communities of inquiry: A review of

182. 167
the literature. Journal of Distance Education, 23(1), 19-48.
Rovai, A. P. (2002). Building a sense of community at a distance. International Review of
Research in Open and Distance Learning, 3(1). Retrieved from
http://www.irrodl.org/index.php/irrodl/article/view/79/153
Russo, T., & Benson, S. (2005). Learning with invisible others: Perceptions of online
presence and their relationship to cognitive and affective learning. Educational
Technology & Society, 8(1), 54-62.
Russo, T. C., & Campbell, S. W. (2004). Perceptions of mediated presence in an
asynchronous online course: Interplay of communication behaviors and medium.
Distance Education, 25(2), 216-232.
Rutter, D. R. (1984). Looking and seeing: The role of visual communication in social
interaction. London: John Wiley.
Rutter, D. R. (1987). Communicating by telephone. Oxford: Pergamon Press.
Rutter, D. R. (1989). The role of cluelessness in social interaction: An examination
of teaching by telephone. In D. Roger & P. Bull (Eds.), Conversation (pp.
294-312). Philadelphia, PA: Multilingual Matters.
Rutter, D. R., Pennington, D. C., Dewey, M. E., & Swain, J. (1984). Eye-contact as a
chance product of individual looking: Implications for the intimacy model of
argyle and dean. Journal of Nonverbal Behavior, 8(4), 250-258.
Ryman, S., Hardham, G., Richardson, B., & Ross, J. (2009). Creating and sustaining
online learning communities: Designed for transformative learning. International
Journal of Pedagogies & Learning, 5(3), 32-45.
Shank, P. (2004). New social interaction tools for online instruction. ITFORUM

183. 168
paper #81. Retrieved from http://it.coe.uga.edu/itforum/paper81/paper81.html
Shea, P., & Bidjerano, T. (2010). Learning presence: Towards a theory of self-efficacy,
self-regulation, and the development of a communities of inquiry in
online and blended learning environments. Computers & Education, 55(4), 1721-
1731.
Short, J. A. (1974). Effects of medium of communication on experimental negotiation.
Human Relations, 27(3), 225-234.
Short, J., Williams, E., & Christie, B. (1976). The social psychology of
telecommunications. London: John Wiley & Sons.
Smith, A. (2010, August 11). Home broadband 2010. Washington, DC: Pew
Research Center’s Internet & American Life Project. Retrieved from
http://pewinternet.org/~/media//Files/Reports/2010/Home%20broadband%20201
0.pdf
So, H. J., & Brush, T. A. (2008). Student perceptions of collaborative learning, social
presence, and satisfaction in a blended learning environment: Relationships and
critical factors. Computers & Education, 51(1), 318-336.
Stacey, E. (2002). Quality online participation: Establishing social presence. Paper
presented at the Research in Distance Education Conference, Deakin University,
Geelong.
Stein, D. S., & Wanstreet, C. E. (2003). Role of social presence, choice of online or face-to-
face group format, and sat with perceived knowledge gained in a distance
learning environment. Paper presented at the Midwest Research to Practice
Conference in Adult, Continuing, and Community Education, Columbus, OH.

184. 169
Steinfield, C. W. (1986). Computer-mediated communication in an organization
setting: Explaining task-related and socioemotional uses. In M. L.
McLaughlin (Ed.), Communication Yearbook 9 (pp. 777-804). Newbury Park,
CA: Sage.
Suler, J. (2004). The online disinhibition effect. CyberPsychology & Behavior, 7(3),
321-326.
Swan, K. (2002). Immediacy, social presence, and asynchronous discussion In J. Bourne
& J. C. Moore (Eds.), Elements of quality online education (Vol. 3, pp. 157-172).
Needham, MA: Sloan Center for Online Education.
Swan, K. (2003). Developing social presence in online course discussions. In S. Naidu
(Ed.), Learning and teaching with technology: Principles and practices (pp. 147-
164). London: Kogan Page.
Swan, K. P., Richardson, J. C., Ice, P., Garrison, R. D., Cleveland-Innes, M.,
& Arbaugh, J. B. (2008). Validating a measurement tool of presence in
online communities of inquiry. e-mentor, 2(24)
Swan, K., Shea, P., Richardson, J., Ice, P., Garrison, D. R., Cleveland-Innes, M., &
Arbaugh, J. B. (2008). Validating a measurement tool of presence in online
communities of inquiry. E-Mentor, 2(24), 1-12. http://www.e-mentor.
edu.pl/e_index.php?numer=24&all=1
Swan, K., & Shih, L. F. (2005). On the nature and development of social presence in
online course discussions. Journal of Asynchronous Learning Networks, 9(3),
115-136.

185. 170
Tallent-Runnels, M. K., Thomas, J. A., Lan, W. Y., Cooper, S., Ahern, T. C., Shaw, S.
M., et al. (2006). Teaching courses online: A review of the research. Review of
Educational Research, 76(1), 93-135.
Thurlow, C., Lengel, L., & Tomic, A. (2004). Computer mediated communication: Social
interaction and the Internet. Thousand Oaks, CA: Sage.
Tu, C.-H. (2000). On-line learning migration: From social learning theory to social
presence theory in a CMC environment. Journal of Network and Computer
Applications, 2, 27-37.
Tu, C.-H. (2001). How Chinese perceive social presence: An examination of interaction
in online learning environment. Education Media International, 38(1), 45-60.
Tu, C.-H. (2002a). The impacts of text-based CMC on online social presence. The
Journal of Interactive Online Learning, 1(2). Retrieved from http://www.
ncolr.org/jiol/issues/PDF/1.2.6.pdf
Tu, C.-H. (2002b). The measurement of social presence in an online learning
environment. International Journal on E-Learning, 1(2), 34-45.
Tu, C.-H., & Corry, M. (2004). Online discussion durations impact online social
presence. In C. Crawford. et al. (Ed.), Proceedings of Society for Information
Technology and Teacher Education International Conference 2004 (pp. 3073-
3077). Chesapeake, VA: AACE.
Tu, C.-H., & McIsaac, M. (2002). The relationship of social presence and interaction in
online classes. The American Journal of Distance Education, 16(3), 131-150.

186. 171
Vrasidas, C., & Glass, G. V. (2002). A conceptual framework for studying distance
education. In C. Vrasidas & G. V. Glass (Eds.), Distance education and
distributed learning (pp. 31-55). Greenwich, CT: Information Age Publishing.
Wagner, E. D. (1994). In support of a functional definition of interaction. The American
Journal of Distance Education, 8(2), 6-29.
Walters, D. (2011, February). The rewrite: Who’d have thought Idaho would be a haven
for radial education reform? The Pacific Northwest Inlander. Retrieved from
http://www.inlander.com/spokane/article-16203-the-rewrite.html
Walther, J. B. (1992). Interpersonal effects in computer-mediated interaction: A
relational perspective. Communication Research, 19, 52-90.
Walther, J. B. (1994). Anticipated ongoing interaction versus channel effects on
relational communication in computer-mediated interaction. Human
Communication Research, 20, 473-501
Walther, J. B. (1996). Computer-mediated communication: Impersonal, interpersonal,
and hyperpersonal interaction. Communication Research, 23(1), 3-43.
Walther, J. B., Anderson, J. F., & Park, D. W. (1994). Interpersonal effects in computer-mediated
interaction: A meta-analysis of social and antisocial communication.
Communication Research, 21(4), 460-487.
Walther, J. B., & Parks, M. R. (2002). Cues filtered out, cues filtered in. In M. L. Knapp
& J. A. Daly (Eds.), Handbook of interpersonal communication (pp. 529-563).
Thousand Oaks, CA: Sage.

187. 172
Watson, H. (2006, April). Governor signs bill establishing rigorous high school
curriculum. Retrieved from http://www.michigan.gov/som/0,1607,7-192-29939-
141369--,00.html
Weedman, J. (1991). Task and non-task functions of a computer conference used in
professional education: A measure of flexibility. International Journal of
Man-Machine Studies, 34, 303-318.
Wheeler, S. (2005). Creating social presence in digital learning environments:
A presence of mind? Paper presented at TAFE Conference (November 11th), Que
ensland, Australia.
White, J. W., & Lowenthal, P.R. (2011). Minority college students and tacit "codes of
power": Developing academic discourses and identities. Review of Higher
Education, 34(2), 283-318.
Wiener, M., & Mehrabian, A. (1968). Language within language: Immediacy, a channel
in verbal communication. New York: Appleton-Century-Crofts.
Willging, P. A., & Johnson, S. D. (2004). Factors that influence students decision to
dropout of online courses. Journal of Asynchronous Learning Environments, 8(4),
105-118.
Williams, E. (1975). Medium or message: Communications medium as a determinant of
interpersonal evaluation. Sociometry, 38(1), 119-130.
Williams, E. (1977). Experimental comparisons of face-to-face and mediated
communication: A review. Psychological Bulletin, 84(5), 963-976.
Williams, E. (1978a). Teleconferencing: Social and psychological factors. Journal of
Communication, 84, 125-131.

188. Williams, E. (1978b). Visual interaction and speech patterns: An extension of previous
173
results. British Journal of Social and Clinical Psychology, 17, 101-102.
Wilson, C., & Williams, E. (1977). Watergate worlds: A naturalistic study of media and
communication. Communication Research, 4(2), 169-178.
Wise, A., Chang, J., Duffy, T., & Del Valle, R. (2004). The effects of teacher social
presence on student satisfaction, engagement, and learning. Journal of
Educational Computing Research, 31(3), 247-271.
Wise, A.F. & Chiu, M. M. (2011, April). The power of a synthesizer role in online
discussion forums: Encouraging midway summaries drives the knowledge
construction process. Paper presented at the Annual Meeting of the American
Education Research Association, April 2011, New Orleans, LA.
Witmer, D. F. (1997). Risky business: Why people feel safe in sexually explicit on-line
communication. Journal of Computer Mediated Communication, 2(4).
Woo, K., Gosper, M., McNeill, M., Pretson, G., Green, D., & Phillips, R. (2008).
Web-based lecture technologies: Blurring the boundaries between face-to-face
and distance learning. Research in Learning Technology, 16(2), 81-93.