Muve your avatar


Published on

AERA 2011, ARVEL SIG session. MUVE Your Avatar presents the results of the March 2010 dissertation work.

Published in: Education
  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide
  • Social networking + simulation + online gaming Space, movement, presence Interact with content embedded in-world Text and audio communication User-generated creation
  • In addition, the study examined the effects of instructor immediacy on student learning outcomes and student co-presence. The main hypothesis was that the use of immediacy behaviors to animate an instructor’s avatar would result in higher perception of instructor immediacy and social presence, and higher perception of student co‑presence and learning outcomes.
  • Heeter (1992) recognises that presence is a subjective experience but divides the concept presence into three dimensions, personal presence, social presence and environmental presence. Personal presence is according to Heeter a measure of the extent to which and the reasons why a person feel as if she/he is in a virtual world. Social presence refers to the extent to which other beings both living and synthetic exist in the virtual world and appear to react to you. Environmental presence refers to the extent to which the environment itself appears to know that you are there and reacts to you. The experiment investigated issues related to using avatars as a presence medium, specifically immediacy behaviors and issues related to mutual awareness.
  • Biocca: deals with the detection and awareness of the copresence of other’s mediated body
  • Respondents were asked to follow a three phase protocol for participation, all of which was completed online at times and on computers chosen by the student: (a) answer a pre-treatment questionnaire that addressed demographics factors, virtual world familiarity, and course content; (b) view one of the four 28‑minute machinimas reflecting the treatment conditions; and (c) complete post treatment questionnaire regarding their perceptions of instructor immediacy behaviors, social presence, and students’ co-presence.
  • view a 28‑minute machinima (treatment condition) post treatment questionnaire (instructor immediacy behaviors, social presence, and students’ co-presence
  • In this order
  • The instructional sessions were scripted and pre‑recorded. Participants were not able to interact with the instructor and the other students. They only viewed the interactions between them. Perceptions and learning outcomes may vary if participants can directly engage with the instructor and their peers.   2.The prerecorded sessions were 28 minutes in length. Study findings might be different if the sessions were longer or closer to a regular lecture length.   3.The experiment was a one shot exposure. Results might be different with semester long exposure.   4.Participants were recruited from two sections of the same undergraduate course. Findings might be different if students were recruited from other disciplines and other degrees.   5.The recording used a bespectacled and bearded Caucasian male avatar to represent the instructor. A different choice of gender, age, race, dress or other variables could produce different results.
  • Descriptive Statistics provided…..
  • There was a statistically significant difference at the p < .05 level: F (3, 113) = 6.5, p = .000 Further analysis revealed some differences between two groups. Post‑hoc Scheffé comparisons indicated that the mean score for Group 1 HiHi ( M = 64.39, SD = 19.39) was significantly different from Group 3 LoHi ( M = 36.64, SD = 22.80, p = .001) and Group 4 LoLo ( M = 45.12, SD = 32.23, p = .032). Post‑hoc Scheffé did not did not register a mean difference between Group 1 HiHi and Group 2 HiLo and between Group 3 LoHi and Group 4 LoLo
  • The researcher hypothesized that participants exposed to high immediacy behaviors displayed by the instructor’s avatar (Groups 1 HiHi and 2 HiLo) would rate the instructor more highly on a measure of social presence than students exposed to low immediacy instructor behaviors (Groups 3 LoHi and 4 LoLo). A one-way ANOVA indicated a significant difference between group means: F (3, 229) = 35.79, p = .000 These results indicate that participants exposed to high immediacy instructor behaviors in the instructor’s avatar did perceive higher instructor social presence than the low immediacy groups
  • The researcher hypothesized that participants exposed to the high immediacy behaviors displayed by the students’ avatars (Group 1 HiHi and Group 3 LoHi) will rate the student-avatars more highly on a measure of co-presence than students exposed to low immediacy students behaviors (Group 2 HiLo and Group 4 LoLo). An ANOVA analysis of the data collected by the Networked Mind Social Presence scale did not provide any significant differences b However, these results do not align with the open‑ended question that asked students to describe the behaviors of avatar students which positively influenced their perception of the avatar students. All four groups reported factors with student high immediacy groups leading with Group 1 HiHi (81%) and Group 3 LoHi (75%), followed by Group 4 LoLo (71.5%) and Group 2 HiLo (63%). These results suggest that student-avatar immediacy behaviors influence perceived student co-presence. Table 19 presents the categories summarizing the students’ descriptions of student‑avatar behaviors that made the students seem real. These findings are congruent with the co-presence literature that argues that the notion of co-presence shares properties with physical presence and implies behavioral engagement and interaction.etween the groups ( F (3, 170) = 1.53, SD = .208)
  • The researcher hypothesized that participants exposed to high immediacy instructor avatar behaviors (Group1 HiHi, and Group 2 HiLo) would outperform students in the low immediacy treatments (Group 3 LoHi and Group 4 LoLo) on this measure of learning outcomes. Learning outcomes were identified as recall and comprehension of the lesson content and were measured with the immediate posttest. Furthermore, the researcher hypothesized that Group 1 (HiHi) would score higher than all other groups. A one‑way analysis of variance demonstrated no significant difference between the groups as measured by the immediate posttest scores ( F (3, 277) = 1.379, p = .249). Negative correlation: as the avatar instructor immediacy increases, posttest scores decrease. This result suggests that avatar immediacy is a distraction to learning
  • Interestingly, participants would note a specific behavior and interpret it differently. For example, while some participants considered the student avatars’ gestures as showing engagement, other participants interpreted those gestures as irrelevant and distracting. Further studies would benefit from evaluating participants’ reaction to student avatars after an orientation session in Second Life. The next question asked participants if they were distracted by the 3D environment in general (see Table 24), and if so to explain. Findings resulted in inconsistencies between the answers above and this question. Further analysis would be recommended at each participant’s level to identify possible variables not considered in this study.
  • Students who viewed the high immediacy instructor avatars (Group1 HiHi and Group 2 HiLo) rated the immediacy behaviors of the instructor-avatars more highly than students who viewed the low immediacy instructor avatars (Group 3 LoHi and Group 4 LoLo). This difference confirms that avatars in Second Life are capable of projecting effective immediacy behaviors and that such behaviors can be perceived by viewers representing undergraduate college students. The correlation of the measures of immediacy behaviors and social presences, ( r = .769, p = .000) indicates that 59% of the variance in the social presence of avatar instructors in this study can be accounted for by students’ perception of instructor immediacy behaviors. Consistent with this finding are students’ reports on how “real” they perceived the instructor to be (Groups 1 HiHi and 2 HiLo > 60% vs. Groups 3 LoHi and 4 LoLo < 25%).
  • Muve your avatar

    1. 1. Sabine Lawless-Reljic, Ed.D AERA 2011 Annual Meeting, ARVEL SIG session New Orleans, Louisiana, April 10, 2011
    2. 2. <ul><li>Introduction </li></ul><ul><ul><li>Communication Dynamics </li></ul></ul><ul><ul><li>Transactional Distance </li></ul></ul><ul><ul><li>Mediated Social Presence </li></ul></ul><ul><ul><li>MultiUser Virtual Environments </li></ul></ul><ul><li>The Study </li></ul><ul><ul><li>Objectives </li></ul></ul><ul><ul><li>Theoretical Framework </li></ul></ul><ul><ul><li>Methods </li></ul></ul><ul><ul><li>Materials </li></ul></ul><ul><ul><li>Results </li></ul></ul><ul><ul><li>Summary </li></ul></ul><ul><ul><li>Significance </li></ul></ul><ul><ul><li>Recommendations </li></ul></ul>
    3. 3. <ul><li>Communication Dynamics </li></ul><ul><li>Transactional Distance </li></ul><ul><li>Mediated Social Presence </li></ul><ul><li>MultiUser Virtual Environments </li></ul>Why MUVE Your Avatar?
    4. 4. The actors have changed @Sabine Lawless-Reljic, 2010
    5. 5. <ul><li>… is not a physical phenomenon. It is an instructional event in a learning situation. </li></ul>Source:
    6. 6. <ul><li>Property of people, not technology </li></ul><ul><li>Moment-to-moment phenomenal state </li></ul><ul><li>Facilitated by a technological representation of another bring </li></ul><ul><li>Varies of the course of a mediated interaction </li></ul><ul><li>From a low level awareness that another being is co-present to more intense sense of the accessibility of psychological modeling of the other’s intentional states (Biocca & Harms, 2004) </li></ul>
    7. 7. The actors have changed How to infuse your technological proxy of your humanity to be perceived by your students as a present and empathic individual?
    8. 8. <ul><li>Sense of presence: MUVEs have the potential to “significantly reduce the subjective feelings of psychological and social distance often experience by distance education participants” (McKerlich, 2007). </li></ul><ul><li>Sense of place: MUVEs are “richly expressive environments that immerse the participant in a setting that includes sound and visual cues, rich textures, and realistic perspective…and vividly create a sense of place” (Johnson & Levine, 2008). </li></ul><ul><li>Sense of power/autonomy: MUVEs allow users “to move around in the virtual world and see it from different angles, to reach into it, grab it, and reshape it” (Rheingold, 1991). </li></ul><ul><li>>>Multiple media in a two-way technology </li></ul>
    9. 9. <ul><li>PURPOSE </li></ul><ul><li>THEORETICAL FRAMEWORK </li></ul><ul><li>METHODS </li></ul><ul><li>MATERIALS </li></ul><ul><li>RESULTS </li></ul><ul><li>SUMMARY </li></ul><ul><li>SIGNIFICANCE </li></ul><ul><li>RECOMMENDATIONS </li></ul>
    10. 10. <ul><li>… to investigate the effects of instructor immediacy behaviors (verbal and nonverbal) in a 3D interactive and immersive virtual environment on student perception of instructor immediacy and perception of instructor social presence. </li></ul>
    11. 11. Research Questions This experimental study explored the following: RQ1: Do immediacy behaviors projected by the avatar-mediated instructor influence instructor immediacy? RQ2: Do immediacy behaviors projected by the avatar-mediated instructor influence instructor social presence? RQ3: Do immediacy behaviors projected by the avatar-mediated students influence perceived students’ co-presence? RQ4: Do immediacy behaviors projected by the avatar-mediated instructor influence learning outcomes? (*) The researcher predicted that projection of instructor immediacy behaviors through an avatar in the context of didactic instruction would positively influence perception of the instructor immediacy, instructor social presence, and student co-presence. * H4 = avatar-based immediacy behaviors would positively influence cognitive learning. Probe not prediction.
    12. 12. Theoretical Framework PRESENCE Short, Williams, & Christie (1976) Social Presence : The degree of salience of other persons in the interaction and the consequent salience (and perceived intimacy and immediacy) of the interpersonal relationships. >>Cognitive synthesis of several factors such as capacity to transmit information about facial expression, direction of looking, posture and nonverbal cues . Heeter (1992) Personal presence : a measure of the extent to which and the reasons why a person feel as if she/he is in a vw. Social presence : degree to which other beings represented in a vw, both living and synthetic, appear to respond to one’s existence or actions. (behaviorism) Environmental presence : VE responds to a person’s actions. Lombard and Ditton (1997) Illusion of non-mediation.
    13. 13. Theoretical Framework, part 2 PRESENCE Witmer and Singer (1998) Subjective experience of being in one place or environments, even as one is physically situated in another. Rourke, Anderson, Archer, and Garrison (1999) Social Presence : The ability of learners to project themselves socially and affectively into a community of inquiry. Russo (2001) Social presence : degree to which a person is perceived to the “real” in a mediated environment and the degree to which they recognize that they are communicating with another human being through a mediating technology. Biocca and Harms (2004): Copresence = sensory awareness of the embodied other + mutual awareness + attentional awareness
    14. 14. Theoretical Framework, part 3 IMMEDIACY Mehrabian (1966): a measure of the psychological distance which a communicator puts between himself and the object of her communication. Andersen (1971): physical approach and avoidance behaviors that may include verbal and nonverbal components. Christophel (1990): perception of physical or psychological closeness between communicators. This study: Measure of the number, combination, and intensity of immediacy behaviors in relevant and appropriate learning event contexts according to modern American presentation conventions Frequency + Intensity + Appropriate use of behaviors
    15. 15. Theoretical Framework, part 4 IMMEDIACY Immediacy behaviors are conveyed by : <ul><li>Verbal cues (Gorham, 1988) </li></ul><ul><li>Call students by name </li></ul><ul><li>Use inclusive pronouns </li></ul><ul><li>Use humor </li></ul><ul><li>Provide feedback </li></ul><ul><li>Use personal examples </li></ul><ul><li>Ask questions </li></ul><ul><li>Refer to class as “our” class </li></ul><ul><li>Praise students’ work, actions... </li></ul><ul><li>Nonverbal cues (Richmond et al, 1987) </li></ul><ul><li>Move in front of/away from desk </li></ul><ul><li>Gesture while talking </li></ul><ul><li>Use variety of tone </li></ul><ul><li>Use variety of expressions </li></ul><ul><li>Smile at students </li></ul><ul><li>Make eye contact </li></ul><ul><li>Move around classroom </li></ul><ul><li>Other proxemics (body lean, openness, orientation, etc.) </li></ul>
    16. 16. Methods <ul><li>The study replicated design elements used by Schutt (2007) and Schutt, Allen, & Laumakis (2009), the key element being the use of pre-recorded teaching scripts in which the instructor immediacy was manipulated to create higher and lower conditions. </li></ul><ul><li>Treatments were recorded in Second Life and posted on Veoh </li></ul><ul><li>Gestures were selected from the standard library and augmented when necessary with scripted gestures from an animation overrider. </li></ul><ul><li>Group 1 High Immediacy Instructor – High Immediacy Students (HiHi) </li></ul><ul><li>Group 2 High Immediacy Instructor – Low Immediacy Students (HiLo) </li></ul><ul><li>Group 3 Low Immediacy Instructor – High Immediacy Students (LoHi) </li></ul><ul><li>Group 4 Low Immediacy Instructor – Low Immediacy Students (LoLo) </li></ul>
    17. 17. Materials <ul><li>Stimulus Materials </li></ul><ul><li>Population and Sample </li></ul><ul><li>Instrumentation </li></ul><ul><li>Data collection </li></ul>
    18. 18. Materials: Stimulus Materials <ul><li>Immediacy Conditions </li></ul><ul><ul><li>Four pre-recorded teaching were produced based on Schutt’s (2007) adaptation of validated instruments developed by: </li></ul></ul><ul><li>Richmond, Gorham, & McCroskey (1987) to measure nonverbal immediacy. </li></ul><ul><li>Examples of nonverbal items </li></ul><ul><li>Sits behind a desk while teaching.* </li></ul><ul><li>Gestures while talking to class. </li></ul><ul><li>Uses monotone/dull voice while talking to class.* </li></ul><ul><li>Looks at the class while talking. </li></ul><ul><li>Smiles at the class as a whole. </li></ul><ul><li>Moves around the classroom while teaching. </li></ul><ul><li>Has a very relaxed body position while talking to the class. </li></ul><ul><li>Gorham (1988) to measure verbal immediacy. </li></ul><ul><li>Examples of verbal items </li></ul><ul><li>Uses personal examples or talks about experiences s/he had outside of class. </li></ul><ul><li>Asks questions or encourages students to talk. </li></ul><ul><li>Uses humor in class. </li></ul><ul><li>Addresses students by name. </li></ul><ul><li>Refers to class as “our” class or what “we” are doing. </li></ul><ul><li>Provides feedback on student work through comments, discussions. </li></ul>* Presumed to be non-immediate, reverse coded for analysis
    19. 19. Materials: Stimulus Materials Automated Immediacy Behaviors in SL Standard gestures library
    20. 20. Materials: Stimulus Materials Controlled Immediacy Behaviors in SL Animation Overrider Head-Up Display & Menu
    21. 21. Materials: Stimulus Materials Snapshot of Instructor Avatar used in Study
    22. 22. Materials: Population and Sample <ul><li>Participants </li></ul><ul><ul><li>Recruited from two 500-seat sections of an introductory course in Psychology in Spring 2008. </li></ul></ul><ul><li>Treatment </li></ul><ul><ul><li>All treatments were pre-recorded machinimas in Second Life and posted on a free video hosting site (Veoh). </li></ul></ul><ul><li>Experiment </li></ul><ul><ul><li>370 students participated, 281 surveys were used for analysis </li></ul></ul><ul><ul><li>Group 1 (HiHi) n = 68 </li></ul></ul><ul><ul><li>Group 2 (HiLo) n = 74 </li></ul></ul><ul><ul><li>Group 3 (LoHi) n = 69 </li></ul></ul><ul><ul><li>Group 4 (LoLo) n = 70 </li></ul></ul>
    23. 23. Materials: Instrumentation <ul><li>Demographic measures (age, gender, ethnicity) </li></ul><ul><li>Online and Virtual World Familiarity (8 items) </li></ul><ul><li>Immediacy (20 verbal & 14 nonverbal items) </li></ul><ul><li>Social presence (10 items) </li></ul><ul><li>Co-Presence (38 items) </li></ul><ul><li>Pre-test of Cognitive Learning (8 items) </li></ul><ul><li>Post-test of Cognitive Learning (8 items ) </li></ul>
    24. 24. Materials: Data Collection <ul><li>Students signed up by responding to an email request sent through their course’s Blackboard and sending their signed consent form. </li></ul><ul><li>2. Students were randomly assigned to one of four online treatment groups. </li></ul><ul><li>Students had one week to complete the assignment, on their own time, on their personal computer. </li></ul>
    25. 25. Materials: Data Collection (cont’d) Participants then completed the following items: Pre-Treatment Questionnaire 1-Demographic questions 2-Online and virtual world familiarity 3-Pretest of cognitive learning items Treatment 4- Pre-recorded 28-minute machinima Post-Treatment Questionnaire 5-Instructor immediacy questionnaire 6-Instructor social presence questionnaire 7-Student co-presence questionnaire 8-Posttest of cognitive learning items.
    26. 26. Limitations <ul><li>Treatments were scripted and pre-recorded. </li></ul><ul><li>The pre-recorded sessions were short (28 minutes). </li></ul><ul><li>The experiment was a “one shot” exposure. </li></ul><ul><li>Participants were recruited from the same pool. </li></ul><ul><li>The avatar chosen to represent the instructor was a bespectacled and bearded Caucasian male. </li></ul>
    27. 27. Results <ul><li>Pre-session Items </li></ul><ul><ul><li>Demography </li></ul></ul><ul><ul><li>Online and virtual worlds familiarity </li></ul></ul><ul><ul><li>Pretest </li></ul></ul><ul><li>Post-session Items </li></ul><ul><ul><li>Immediacy </li></ul></ul><ul><ul><li>Social presence </li></ul></ul><ul><ul><li>Co-presence </li></ul></ul><ul><ul><li>Cognitive learning </li></ul></ul>
    28. 28. Results: Demography <ul><ul><li>Females 63.7% (n = 179) </li></ul></ul><ul><ul><li>Males 36.3% (n = 102) </li></ul></ul><ul><ul><li>68.7% of participants were 18-19 years old. </li></ul></ul><ul><ul><li>29.6% of participants were 20-26 years old. </li></ul></ul><ul><ul><li>Self-Identified Ethnicity </li></ul></ul>Frequency Percentage White Mexican American Asian/SE Asian Filipino Other Hispanic Other/Not Stated African American International Pacific Islander 118 46 40 25 16 16 12 6 2 42.0 16.4 14.2 8.9 5.7 5.7 4.3 2.1 0.7
    29. 29. Results: Online and Virtual World Familiarity <ul><li>Online Participation </li></ul><ul><ul><li>68.7% ( n = 193) indicated 0-5 hours per week </li></ul></ul><ul><ul><li>67.3% ( n = 189) had never taken courses in which the instructor used online conferencing tools. </li></ul></ul><ul><ul><li>75.1% ( n = 211) had never been in a virtual world. </li></ul></ul><ul><li>Indicated Types of Participation in Online Communities </li></ul>Online Communities Population Percentage Study Hobby Social network Forum Role-Play/Game 141 100 69 58 44 50.2 35.6 24.6 20.6 15.7
    30. 30. Results: Immediacy RQ1: Do immediacy behaviors projected by the instructor-avatar influence instructor immediacy? Mean Scores for Immediacy Significant difference with only pairs: 1 (HiHi) and 3 (LoHi) p = .001 1 (HiHi) and 4 (LoLo) p = .032 Groups Mean Scores Group 1 (HiHi) n = 28 Group 2 (HiLo) n = 29 Group 3 (LoHi) n = 28 Group 4 (LoLo) n = 32 64.39 53.44 36.64 45.13
    31. 31. Results: Social Presence RQ2: Do immediacy behaviors projected by the instructor-avatar influence instructor social presence? Mean Scores for Instructor Social Presence Only pair NOT sig. dif: Group 3 (LoHi) & Group 4 (LoLo) p = .995 Strong positive correlation between avatar immediacy and instructor social presence: r = .769, n = 111, p < .0005 Groups Mean Scores Group 1 (HiHi) n = 62 Group 2 (HiLo) n = 52 Group 3 (LoHi) n = 58 Group 4 (LoLo) n = 61 29.79 20.36 14.15 13.68
    32. 32. Results: Student Co-Presence RQ3: Do immediacy behaviors projected by the student-avatar influence perceived students’ co-presence? Mean Scores for Students Co-Presence Open-Ended Item # 5: Describe the student-avatar behaviors which positively influenced your perception of the students. No sig. dif . Groups Mean Scores Group 1 (HiHi) n = 32 Group 2 (HiLo) n = 50 Group 3 (LoHi) n = 43 Group 4 (LoLo) n = 46 29.53 30.56 21.53 28.71 Aspects of the machinima Number of responses Interactive participation Acted like real students Were attentive Used gestures 45 39 28 22
    33. 33. Results: Learning RQ4: Do immediacy behaviors projected by the instructor-avatar influence cognitive learning outcomes? Mean Scores for Pretest and Posttest no sig. dif. between the groups as measured by the immediate posttest scores ( F (3, 277) = 1.379, p = .249 Negative relationship between avatar immediacy and learning outcomes r = ‑.50, n = 117, p = .000 Groups Pretest Mean Scores Posttest Mean Scores Group 1 (HiHi) n = 68 Group 2 (HiLo) n = 74 Group 3 (LoHi) n = 69 Group 4 (LoLo) n = 70 3.67 3.75 4.14 3.50 5.57 5.02 5.63 5.35
    34. 34. Results: Open-Ended Items Item 1: Number of students who indicated they perceived the instructor avatar as real Item 2: categories of Aspects of the machinima that made the instructor avatar seem real Perceived Instructor as ‘real’ Group 1 (HiHi) n = 65 Group 2 (HiLo) n = 67 Group 3 (LoHi) n = 67 Group 4 (LoLo) n = 64 Yes No 53 12 41 23 16 51 18 46 Aspect of the machinima Number of responses It was a real voice Used gestures, facial expressions, moved body His character looked like a real person/teacher He was accessible (personal info, caring, ‘us’, emotions…) Interacted with students, called them by first names Encouraged students to be involved, asked questions 79 44 32 29 27 21
    35. 35. Results: Open Ended Items Items 5 & 6: Positive/Negative Student-Avatars’ Behaviors Item 7: Environment is distracting Positive Behaviors Negative Behaviors Participated in class Felt together like a classroom Used humor Were different and unique Interacted with instructor Were friendly Sitting attentive Were like real students Move realistically Showed engagement Did not do anything Looked bored It’s fake anyway Diversity is distracting Ridiculous, weird Slowed down instructor Did not look realistic Did not talk Gestures distracted They can’t see me Group 1 ( n = 64) Group 2 ( n = 66) Group 3 ( n = 60) Group 4 ( n = 66) Yes No 24 40 31 35 31 35 15 41
    36. 36. Summary <ul><li>Instructor-avatar immediacy behaviors influence instructor immediacy: Group 1 HiHi and Group 2 HiLo > Group 3 LoHi and Group 4 LoLo </li></ul><ul><li>Instructor-avatar immediacy behaviors influence instructor social presence: 59% of the variance in the social presence of avatar instructors can be accounted for by students’ perception of instructor immediacy behavior. </li></ul><ul><li>Student-avatars immediacy behaviors influence perceived student co-presence: The networked minds measure results showed no sig. relationship ≠ participants reported awareness of student-avatars activities. </li></ul><ul><li>All groups improved pre to post for cognitive learning. Inconclusive. </li></ul>
    37. 37. Recommendations <ul><li>Paralanguage behaviors </li></ul><ul><li>Voice quality </li></ul><ul><li>Emotion & speaking style </li></ul><ul><li>Prosodic features </li></ul><ul><li>NonVerbal </li></ul><ul><li>Proxemic behavior, body language/posture </li></ul><ul><li>Facial expression & gaze direction </li></ul><ul><li>Gestures </li></ul><ul><li>Text </li></ul><ul><li>Tone, style </li></ul><ul><li>Spatial arrangement of words </li></ul><ul><li>Use of emoticons </li></ul><ul><li>Use of symbols and infographics </li></ul><ul><li>Avatar </li></ul><ul><li>Clothing, hairstyles </li></ul><ul><li>Physique </li></ul>