Vol 15 No 5 - April 2016

International Journal
of
Learning, Teaching
And
Educational Research
p-ISSN:1694-2493
e-ISSN:1694-2116IJLTER.ORG
Vol.15 No.5

PUBLISHER
London Consulting Ltd
District of Flacq
Republic of Mauritius
www.ijlter.org
Chief Editor
Dr. Antonio Silva Sprock, Universidad Central de
Venezuela, Venezuela, Bolivarian Republic of
Editorial Board
Prof. Cecilia Junio Sabio
Prof. Judith Serah K. Achoka
Prof. Mojeed Kolawole Akinsola
Dr Jonathan Glazzard
Dr Marius Costel Esi
Dr Katarzyna Peoples
Dr Christopher David Thompson
Dr Arif Sikander
Dr Jelena Zascerinska
Dr Gabor Kiss
Dr Trish Julie Rooney
Dr Esteban Vázquez-Cano
Dr Barry Chametzky
Dr Giorgio Poletti
Dr Chi Man Tsui
Dr Alexander Franco
Dr Habil Beata Stachowiak
Dr Afsaneh Sharif
Dr Ronel Callaghan
Dr Haim Shaked
Dr Edith Uzoma Umeh
Dr Amel Thafer Alshehry
Dr Gail Dianna Caruth
Dr Menelaos Emmanouel Sarris
Dr Anabelie Villa Valdez
Dr Özcan Özyurt
Assistant Professor Dr Selma Kara
Associate Professor Dr Habila Elisha Zuya
International Journal of Learning, Teaching and
Educational Research
The International Journal of Learning, Teaching
and Educational Research is an open-access
journal which has been established for the dis-
semination of state-of-the-art knowledge in the
field of education, learning and teaching. IJLTER
welcomes research articles from academics, ed-
ucators, teachers, trainers and other practition-
ers on all aspects of education to publish high
quality peer-reviewed papers. Papers for publi-
cation in the International Journal of Learning,
Teaching and Educational Research are selected
through precise peer-review to ensure quality,
originality, appropriateness, significance and
readability. Authors are solicited to contribute
to this journal by submitting articles that illus-
trate research results, projects, original surveys
and case studies that describe significant ad-
vances in the fields of education, training, e-
learning, etc. Authors are invited to submit pa-
pers to this journal through the ONLINE submis-
sion system. Submissions must be original and
should not have been published previously or
be under consideration for publication while
being evaluated by IJLTER.

VOLUME 15 NUMBER 5 April 2016
Table of Contents
Promoting Geoscience STEM Interest in Native American Students: GIS, Geovisualization, and
Reconceptualizing Spatial Thinking Skills ..........................................................................................................................1
Donna M. Delparte, R. Thomas Richardson, Karla Bradley Eitel, Sammy Matsaw Jr. and Teresa Cohn
Using Coh-Metrix to Analyze Chinese ESL Learners’ Writing....................................................................................... 16
Weiwei Xu and Ming Liu
The Factors Affecting the Adaptation of Junior High School Students with Severe Disabilities to Inclusive or
Segregated Educational Settings ........................................................................................................................................ 27
Li Ju Chen
Supporting to Learn Calculus Through E-test with Feedback and Self-regulation .................................................... 43
Yung-Ling Lai and Jung-Chih Chen
Authentic Instructional Materials and the Communicative Language Teaching Approach of German as Foreign
Language in Uganda ............................................................................................................................................................ 61
Christopher B. Mugimu and Samuel Sekiziyivu
An Evaluation of the New School Administrator Assignment System Applied in Recent Years in Turkey............ 75
Tarık SOYDAN
Antecedents of Newly Qualified Teachers’ Turnover Intentions: Evidence from Sweden ...................................... 103
Dijana Tiplic, Eli Lejonberg and Eyvind Elstad
Multiple Intelligences in the Omani EFL context: How Well Aligned are Textbooks to Students’ Intelligence
Profiles? ............................................................................................................................................................................... 128
Fawzia Al Seyabi and Hind A’Zaabi

1
©2016 The authors and IJLTER.ORG. All rights reserved.
International Journal of Learning, Teaching and Educational Research
Vol. 15, No. 5, pp. 1-15, April 2016
Promoting Geoscience STEM Interest in Native
American Students: GIS, Geovisualization, and
Reconceptualizing Spatial Thinking Skills
Donna M. Delparte and R. Thomas Richardson
Idaho State University
Pocatello, Idaho, USA
Karla Bradley Eitel, Sammy Matsaw Jr. and Teresa Cohn
University of Idaho
Moscow, Idaho, USA
Abstract. Recent innovations in Geographic Information Systems (GIS)
and geovisualization tools offer new opportunities to promoting interest
in geoscience and STEM careers with Native American Students. The
place-based educational model is particularly suited to geoscience
education and can appeal to Native American students’ connection to
local places. Yet the geoscience discipline is heavily imbued with
Western Science conceptions of places, spaces, and physical processes
that are not in congruence with the interconnected worldview of
Indigenous Science. This review of the literature on geoscience
education offers three recommendations to promote geoscience and
STEM interest among Native American youth. The practice of science is
a field that has only been recently contested by the Indigenous Science
worldview. This cognitive dissonance between Native American
students who have a deep attachment to their local environment can be
at odds with the objective perspective of Western science. The place-
based educational model aligns with Indigenous Science and prior
research has shown that it promotes STEM and geoscience in Native
American students. Since GIS and geovisualization tools are well-suited
to place based education and promote spatial thinking skills, which
have been identified as crucial to geoscience and STEM success, this
review provides several examples of research and education projects
using these technologies. Yet our understanding of spatial thinking is
based on Western Science’s conceptions of space as an abstract quality.
We contend that like other areas of science which are increasingly more
open to Indigenous Science practices, spatial thinking research needs to
do likewise by developing an analytical framework that accommodates
Native American ideas on space and place. We draw on recent research
to frame an argument for advancing research on creating an interwoven,

2
hybrid conception of spatial thinking that can accommodate both
Western and Indigenous Science perspectives.
Keywords: Geoscience education; Native American students;
geovisualization; place-based education; spatial thinking.
1. Introduction: Defining the Issue
This review of the literature on promoting STEM (Science, Technology,
Engineering, and Mathematics) geoscience interest among Native American students has
three objectives. The first is to propose connections between the practices of Indigenous
Science and place-based learning. We begin with a discussion of Indigenous Science and
its emphasis on place as distinct from the Western Science tradition, which may dissuade
some Native American students from pursuing STEM education. Similarities between
Indigenous Science perspectives and place-based educational practices are then
compared to establish a common ground for identifying several programs that have
successfully integrated these two approaches to engage students in geoscience learning.
The second objective is to extend this connection by describing several recent
innovations in GIS (Geographic Information Systems) and geographic visualization that
apply to developing spatial thinking skills— an important element of STEM competency
in geoscience education and careers. These technologies are readily adaptable to place-
based learning and can enable all students to understand their local spaces in new ways
by developing their spatial thinking skills. Yet the literature on fostering STEM interest
among Native American students is sparse in terms of connecting spatial thinking
strategies to geoscience programs. Hence, our final objective is to propose means to
redefine space and place that are compatible with both the Indigenous Science and
Western Science. We capitalize upon the geographic construct of respatialization to
frame a proposal for further research and debate between the cognitive science,
geoscience, and Indigenous science.
2. Western and Indigenous Science: Issues of Space and Place
Native Americans have a rich and deep attachment to locale, especially within
their traditional homelands; it is the source of their cultural traditions and knowledge
(Cajete, 1994; 2000). Culture and history thus influence their conceptions of natural
events, where humanity is part of the natural world (Cajete, 2000; Semken, 2005).
Embedded within Native culture across North America, a strong sense of place is
evident; space is both culturally constructed and highly localized (Cajete, 1994, 2000;
Doering & Veletsianos, 2008; Semken, 2005). Therefore, spatial awareness (rather than
spatial thinking, per se) is of particular, embodied importance to many Native
Americans. This is a tradition that is dichotomous with the norms of Western Science’s
idea of space as an abstract set of Cartesian coordinates whereby the human and natural
environments are separated. Spatial thinking, as seen by Western Science, may be
perceived as reductionist in comparison to a more holistic sense of space and place
within Native American cultural traditions. This dichotomy between Western and
Native American perspectives on space has been expressed through hegemonic Western
cartographic practices (Harvey, 1984: Palmer, 2012); maps are used to categorize space in
non-Native terms. It is not surprising that studies have documented that American
Indian students, like other ethnic or racial minorities, are underrepresented in
geoscience education (Riggs & Semken, 2001; Semken, 2005) and STEM education in
general (Babco, 2003; Wang, 2013).

3
A growing body of research promotes Indigenous Science as a culturally
responsive alternative to Western Science (Cajete, 1999; Castagno & Brayboy, 2008;
Snively & Corsiglia, 2001).Western Science represents a divergent, even oppositional,
view of space and place while Indigenous Science navigates space both synchronously
and without division (Cajete, 1999). Some perceive the emergence of Indigenous Science
(Cajete, 1999; Snively & Corsiglia, 2001) as a reaction to the hegemonic power and
authority of Euro American culture. Researchers have identified ―some form of cultural
discontinuity as a root cause‖ (Semken, 2005, p. 150), which may disempower Native
American students from pursuing STEM education and careers; they must separate the
cultures of their daily lives within the culture of Western Science (Aikenhead, 1998).
Although it is important to acknowledge that Indigenous Science represents an
array of relationships and experiences, there is no singular conception of Indigenous
Science (Castagno & Brayboy, 2008). Nevertheless, some generalizations may be made. It
includes multiple way of knowing that are based on interaction with the local
environment (Cajete, 2009). Thus, Indigenous Science generally supports holism rather
than reductionism (Cajete 2000) and subjectivity over objectivity (Cajete, 1999); the
intertwining of physical and spiritual aspects of the universe (Castagno and Brayboy,
2008); and a person relationship between people and their environment (Deloria, 2003).
Reconciling the divergent analytical lenses of Western and Indigenous Science
may allow culture, knowledge, and place to be more interconnected, thereby promoting
more STEM engagement among Native American students. Our proposition is more
limited in scope. We will focus on how the similarities between place-based education
and Indigenous Science are articulated to boost STEM interest in geoscience learning. We
will then examine how innovations in spatial thinking, as enabled by new geoscience
visualization tools, can be used to foster STEM interest in Native American students.
Emerging from this discussion, suggestions to re-conceptualize the underlying processes
of spatial thinking in geoscience education will be proposed for future research and
dialogue.
3. Place-Based Learning and Indigenous Science
Place-based education, like Indigenous Science, utilizes a holistic,
engaged approach to understanding processes and relationships. Sobel (2004)
defined place-based education as
the process of using the local community and environment as a starting
point to teach concepts in language arts, mathematics, social studies,
science, and other subjects across the curriculum. Emphasizing hands-
on, real-world learning experiences, this approach to education
increases academic achievement, helps students develop stronger ties to
the community, enhances students’ appreciation for the natural world,
and creates a heightened commitment to serving as active, contributing
citizens. Community vitality and environmental quality are improved
through active engagement of local citizens, community organizations,
and environmental resources (p. 7).
A central characteristic and distinguishing feature of place-based education is to
break down artificial constructs and barriers such as distinctions between school,
community, nature, and humanity. Geoscience education contributes to place-based
education (Apple, Lemus & Semken, 2014). Semken (2005) identified five characteristics
of place-based geoscience education: (1) content focusses on the geological
characteristics of particular locales from an Earth systems perspective; (2) recognize and
validate that places have varied meanings for different groups; (3) hands-on, authentic

4
research occurs in the locale and is taught by and shared with those who live there; (4)
research efforts and results respect environmental and cultural sustainability; and (5)
teaching goals are to build a shared attachment to a place amongst students, instructors,
and researchers and can have the indirect benefit of promoting STEM engagement. A
recent study examining a place-based instructional model to teach geoscience in an
urban environment reported an increase in student science interest (DeFelice, Adams,
Branco, & Pieroni, 2013). Likewise, positive results with respect to place-attachment have
been reported with respect to indigenous-oriented geology courses (Semken & Freeman,
2008; Johnson, Sievert, Durglo, Finley, Adams & Hoffman, 2014), which may play a
factor in Native American students’ STEM interest.
Many theorists perceive strong relationships between Indigenous Science
and place-based education (Apple et al., 2014; Semken, 2005; Semken &
Freeman, 2008; Semken, Neakrase, Dial, & Baker, 2009; Zalles, Collins,
Montgomery, Colonesese & Updegrave, 2005). Place-based education ―is
advocated as a way to improve engagement and retention of students,
particularly members of indigenous or historically inhabited communities (e.g.,
American Indian, Alaska Native, Native Hawaiian, Mexican American) who
possess rich culturally-rooted senses of the places studied‖ (Semken & Freeman,
2008, p. 1044). The place-based education model represents a critical
reinterpretation of Western education. Place-based learning is holistic, situated,
and opposed to globalization because of its emphasis on environmental and
socio-cultural sustainability. The constructivist learning modalities used in
place-based education include experiential learning, problem-based teaching
approaches, interdisciplinary focus on content delivery, peer teaching,
recognition of students’ unique abilities, and environmental awareness and
appreciation. Place-based education is a particularly useful educational
philosophy for engaging with Native American students because of its focus on
sense of place, community engagement, and holistic learning that uses creative
expression, as well as scientific observation, in studies of place. Rather than
basing itself in a cultural framework, however, place-based education uses local
environments and communities to teach an integrated curriculum (Sobel, 2004),
so it may lack the linguistic and cultural elements of many Native American-
specific traditional knowledge programs.
Place-based education in many Native communities is realized through formal
contexts via indigenous language immersion schools, such as the Aha Punana Leo
programs in Hawaii; Cuts Wood School of the Blackfeet Nation; Waadookodaading, the
Ojibwe Language Immersion Charter School; and the Nikaitchuat Ilisagviat immersion
school of the Qukiktagrukmiut people. In these language immersion programs, the
language does not make sense unless the place you inhabit becomes a part of you and
you a part of it. Because of this, it makes sense that place-based education’s formal and
informal learning contexts agree with indigenous ways of thinking and communicating.
Other programs sponsored through school districts, such as the North Vancouver School
District’s Aboriginal Education Program, also provide opportunities for Native and non-
native students to learn Coast Salish traditions and practices within a place-based
learning milieu. Since place-based and Indigenous Science practices share numerous
attributes, the following sections will discuss their common strategies to promote STEM
interest.
4. Indigenous Science, Place-Based Education, and STEM

5
Place-based education has been linked to STEM interest. A meta-analysis of the
efficacy of place-based teaching in 40 U.S. schools found evidence for increased scientific
knowledge, reading, writing math and social studies scores, compared to traditional
science course teaching methods using standardized test scores (Lieberman & Hoody,
1998). There is a growing body of evidence connecting place-based geoscience education
and Indigenous Science that promotes STEM engagement (Adetunji, Ba, Ghebreab,
Joseph, Mayer, & Levine, 2012; Morgan & Semken, 1997; Semken, 2005; Semken &
Freeman, 2008; Semken, Freeman, Watts, Neakrase, Dial, & Baker, 2009). This review
will focus on several examples of geoscience courses which promote spatial thinking and
awareness.
Locally-driven, place-based educational programs may offer a viable option for
Native American students that is more culturally-sensitive. Geoscience courses, with
their emphasis on place and space, can be relevant to Native American students and
thus serve as a gateway to further STEM interest. Collaborative efforts to encourage
Native American students to enter geoscience careers have been advanced through the
Indigenous Earth Sciences Project and the Sharing the Land Program and, like similar
place-based educational initiatives, have increased the number of Native American
geoscientists who can apply their expertise in local communities (Riggs, Robbins, &
Danner, 2007). There are numerous studies of place-based geoscience instructional
programs reporting an increase in student science interest: the Geosciences Awareness
Program (Adetunji et al., 2012) and a study of geoscience learning in urban parks ;
(DeFelice, Adams, Branco, & Pieroni, 2013) are two recent examples. Zalles et al. (2005)
implemented a project to foster STEM interest in a fluvial geomorphology course for
high school and undergraduate students (noteworthy here because a large number of
Native Americans participated in the study). Although the results of the high school
course were inconclusive, increases in STEM interest and attachment to place were
statistically significant in the undergraduate course based on a Science Motivation
Questionnaire (SMQ) and Place Attachment Inventory (PAI). Similarly, the PAI and the
Place Meaning Survey (PMS) indicated a statistically significant increase in identification
and attachment to place in a pre- and post-survey of undergraduate students in an
indigenous knowledge geoscience course at the University of Arizona (Semken &
Freeman, 2008). Similar instruments by Shamai (1991), Kaltenborn (1998), and Williams
and Vaske (2003) were used in a variety of studies to also measure place
identity/attachment.
There are two studies most pertinent to the argument we will propose to
redefine spatial thinking. Tsé na’alkaah, an Indigenous Physical Geology course offered
at Arizona State University conceptualizes environmental change as interactions
between the Earth (Nohosdzáán) and Sky (Yáhdilhil) and are interweaved in the stories
of Navajo tribes living within the area of study (Morgan & Semken, 1997; Semken, 2005).
Western Science terms were given Navajo labels to develop a sense of place imbued with
personal meaning. Their resulting Earth systems framework represents a hybrid of
Indigenous and Western Science knowledge. Likewise, Palmer (2012) explored the use of
the Kiowa language for spatial labels and concepts represented in an indigital
geographic information network (iGIN), a ―synthesis of indigenous and scientific spatial
knowledge‖ (p. 81). Both Semken and Palmer recognized the importance of language as
a carrier of cultural meaning regarding spatial terms and concepts that have traditionally
been co-opted by Western cognitive spatial science practices. For example, Palmer noted
a lack of research on incorporating Indigenous languages into GIS analysis. We propose
extending the Semken and Palmer’s lines of research by advocating for a holistic way of
thinking spatially by recapturing the language of space and place—an issue that will be
addressed in the final section of this review.
In spite of these examples, many place-based educational programs are located
outside of Native communities and, therefore, inaccessible to many Native students

6
(Semken, 2005). Complicating this is a shortage of mentors and science role models,
inadequate teaching facilities, under-trained teachers (Syed, Goza, Chemers &
Zurbriggen, 2012), as well as an absence of earth science courses beyond the middle-
school level in many states, perhaps another factor that reflects low completion rates of
STEM degrees in tribal colleges (Babco, 2003). Furthermore, traditional science curricula
and textbooks tend to present a linear, mechanistic, and process-driven view of
environmental systems, which runs contrary to the Native American understanding of
the non-linear, cyclical understanding of environmental interactions (Semken, .
Adolescents who experience STEM-related discrimination or stereotyping within the
structural power and knowledge relations inherent to public education may question
their own abilities or compatibility with STEM study and therefore may be reluctant to
explore or pursue these areas (Grossman & Porche, 2013). Yet recent advances in
managing and viewing geoscience data can offer new ideas on teaching and learning
that have the potential to engage students and teachers from western and non-western
pedagogies in new ways.
5. Geographic Information Systems (GIS) and Geovisualization as
Tools to Promote STEM Interest
Two specific geoscience technologies are offered as potential means to
promote STEM interest in Native American students through place-based
instruction: geoscience education using GIS and geovisualization tools to
promote spatial thinking. These two examples empower learners to explore
their own localities while developing scientific thinking and skills that will
benefit them in further STEM education and careers. Although these strategies
may have broad appeal to other student populations, testing their educational
effectiveness in rural Native American communities must entail a collaborative
partnership between educators, researchers, and tribal members. A number of
past and current research initiatives will be highlighted as examples of these
technologies in action.
5.1. Geoscience education and GIS as spatial learning tools.
Geospatial learning increases higher order cognitive thinking and
engages students to use geospatial data to construct their own interpretation of
places and spaces, which is consistent with both experiential and constructivist
educational theory (Doering & Veletsianos, 2007). The ability to think in spatial
terms is considered to be a key skill that is ―universal and useful in a wide
variety of academic discipline and everyday problem-solving situations‖ (Lee &
Bednarz, 2009, p. 183). The National Academy of Sciences (Downs & DeSouza,
2006), in their study Support for Thinking Spatially: The Incorporation of Geographic
Information Science Across the K-12 Curriculum, regarded spatial thinking to be on
par with mathematical and verbal thinking skills. Since spatial literacy is a
newly recognized area of knowledge, it is an avenue worthy of additional
evidence-based research (Bednarz, 2004; Schulz, Kerski & Patterson, 2008) and
using it to understand locales makes it a natural fit for place-based educational
programs.
GIS is a useful tool for visualizing the interrelationships of spatial attributes.
Studies on the use of GIS as an educational platform in public schools have
demonstrated its potential to increase spatial thinking (Doering & Veletsianos, 2007; Lee

7
& Bednarz, 2009), as well as positive attitudes to science and technology (Baker & White,
2003). Studies on the use of GIS applications such as Google Earth and ArcGIS Explorer
(Doering & Veletsianos, 2008; Lee & Bednarz, 2009) and Virtual Globe (Schultz, Kerski &
Patterson, 2008) provide examples of successful, evidence-driven applications of existing
GIS tools for teaching both at the K-12 and college levels. GIS applications are, by their
very nature, embedded with web functionality, and are also suitable for advancing
spatial thinking and geographic knowledge in e-learning environments (Lynch, Bednarz,
Boxall, Chalmers, France & Kesby, 2008).
The Western Consortium for Water Analysis, Visualization and Exploration
program (WC-WAVE) sponsors the Undergraduate Visualization & Modeling Network
Program (UVMN), a training forum for undergraduate students and supporting faculty
at regional colleges in Idaho, Nevada, and New Mexico. Native American geoscience
students are included in this program and have an opportunity to work collaboratively
on GIS-enabled place-based studies and use novel techniques for visualization and data
exploration (National Science Foundation award # IA-1301346).
5.2. New frontiers: geovisualization.
Geovisualization takes geographic data, usually from a GIS database,
and converts it into interactive and predictive three dimensional models that
enables spatial relationships to be viewed in innovative ways (Kinzel, 2009;
Kraak, 2003; MacEachren & Kraak, 1997; 2001). Geovisualizations are
constructed from highly-detailed spatial models. For example, a digital
elevation model (DEM) can be used as a template over which spatial data
(Google Earth or other remotely-sensed or field collected GPS data) is layered.
Geovisualizations need not be restricted to maps, however. Photos and
video, for example, can be integrated with spatial data in a geovisualization. 360º
gigapans allow for the creation of virtual tours based on digitally stitched set of
photos that facilitate explorations into sites of interest. This panoramic image of
a landscape can contain links to information associated with parts of the image.
Links can display text information, hyperlinks, other maps, video, or reorient the
viewer to a new location. An example of a gigapan virtual tour was created for
the Portneuf River to facilitate re-visioning by the city of Pocatello through the
Managing Idaho’s Landscape through Ecosystem Services (MILES) project. It
can be viewed at the following address:
http://miles.isu.edu/Greenway/Greenway.html. Another example from the
MILES project is a 3-D future urban redevelopment along the Portneuf River
channel using ESRI’s City Engine This may be viewed online at
http://miles.isu.edu/visualizations.shtml3-D. In this example, objects from
mapped data can be associated with rules and attributes and map layers can be
toggled to visualize patterns and associations between spatial data. The
visualization scenarios for city planning are integrated into Unity 3D and
provide an immersive environment on a virtual reality headset (Delparte,
Johnson & Tracy, in prep).
Other technologies can also be used to enhance digital maps. For
example, photographs from inexpensive digital cameras can be stitched together
using Structure from Motion (SfM) technology to create 3-D models (Bolles,
Baker, Marimont, 1987; Koenderink & Van Doorn, 1991). Sketchfab software
allows for online storage of images. Microsoft’s Kinect sensor can also create 3D
scans and are being used to catalogue native artifacts for remote online viewing

8
(Youngs & Delparte, in prep). Interactive and predictive 3D models can be
created to identify spatial relationships in large geographic datasets. This
technology has built predictive 3-D maps of a Native Hawai’ian cultural
landscape for environmental monitoring and preservation of Hawai’i’s Lake
Waiau (Delparte, Belt, Nishioka, Turner, Richardson & Ericksen, 2014) and coral
reef fisheries in the Northwest Hawai’ian islands (Burns, Delparte, Gates &
Takabayashi, 2015).
Geovisualization platforms coupled to current research on GIS, mobile
computing, and pedagogy have the potential to increase student engagement
and learning. Cost need not be a significant obstacle to using these technologies.
Everyday tools (such as cameras, tablets, and cell phones) can be used to capture
scientific data and much of the image processing software is freely available
online (i.e. 123D Catch is a freeware photo stitching tool that can transform a
series of photographs into 3D models).
Recent studies have demonstrated the educational benefits of using
geovisualization tools to teach spatial thinking skills (Hauptman & Cohen, 2011;
Lee & Bednarz, 2009; Kinzel, 2009; Schultz et al., 2008, Titus & Horsman, 1996),
yet additional research on the linkages between visual and spatial thinking and
how they can be promoted through geovisualization is needed (Kinzel, 2009;
Montello, 2009; Vogler, Ahamer, & Jekel, 2010). The authors are collecting
evidence-based research to examine the specific learning benefits and measures
of cognitive load associated with the use of geovisualization technologies
(Richardson, in prep). An example is an upcoming study comparing learning
performance and cognitive load of two dimensional, three dimensional, and
tactile feedback geovisualization maps. The goal is to select the most appropriate
interface for teaching spatial thinking using maps and to offer suggestions for
designing instructional programs that promote spatial cognitive processing in
learning.
Geospatial technologies have real-world relevance for jobs that are
meaningful to Native students and can thus enhance STEM interest. For
example, natural resource professionals working for the Shoshone-Bannock
Tribes in southeast Idaho use ArcGIS Collector on iPads to sample biological
data in the field. Many tribes hire GIS professionals to advise and inform natural
resource management departments. Therefore, geospatial tools may encourage
students to build skills that eventually allow them to find a career within their
tribal community, using tools that convey a multiplicity of perceptions in
―symbols of place‖ (Cajete, 1999), a theme that will be explored in the next
section.
Semken (2005) criticized geoscience for instruction that ―emphasizes global
syntheses over exploration and in-depth understanding of places that have prior
meaning for Indigenous students, and may even depict such places in culturally-
inappropriate ways‖ (p. 149). We do not deny that geoscience has and can promulgate
Western Science thinking to the detriment of other perspectives. In the final section, we
propose an alternative approach that respects a dualistic understanding of space and
place through geovisualization.
6. Expanding Our Understanding of Spatial Thinking to Incorporate
Multiple Perspectives

9
Recent research on what constitutes a definition for and characteristics of
spatial thinking, from a Western perspective, are rooted in the cognitive
sciences. Spatial thinking refers to the cognitive aspects of (1) visualizing and
recalling spatial information such as shape, dimension, relative location, or
perspective and (2) mentally representing and manipulating objects that are
either in a two dimensional or three dimensional format (Downs & DeSouza,
2006; Velez, Silver & Tremaine, 2005). Some researchers consider spatial thinking
as distinct from other more generic terms as kinesthetic ability or spatial
awareness (Fleishman & Rich, 1963).
There has been a substantial body of research in the realm of spatial thinking as
an important, yet overlooked area of skill and knowledge in K-12 American education
(Downs & DeSouza, 2006). Numerous studies have further examined spatial thinking
and its relationship to Native American learning preferences (Apple et al., 2014; Cajete,
1994, 2000; Bednarz, 2004; Pewewardy, 2002; Semken, 2005). These characteristics
include a strong social emphasis, holistic learning, creative expression, respect for
cultural traditions, and use of story-telling as an effective medium for delivering
knowledge (Pewewardy, 2002). Cajete (1999) recommended less emphasis on verbal
learning, preferring kinesthetic, spatial, and visual learning activities and understanding
processes from examples, which then lead to abstract concept formation.
Recent studies have been devoted to categorizing the components of
spatial thinking (Bednarz & Lee, 2011; Gersmehl & Gersmehl, 2006). A taxonomy
of spatial thinking skills have been proposed: defining a location; describing
conditions; tracing spatial connections, making spatial comparison; inferring a
spatial aura; delimiting a region; fitting a place into a spatial hierarchy; graphing
a spatial transition; identifying a spatial analog; discerning spatial patterns;
assessing a spatial association; designing and using a spatial model; and
mapping spatial exceptions (Gersmehl & Gersmehl, 2006). From a
geographic/cartographic perspective, and couched in the language of Western
Science, these characteristics labels are reasonable, particularly when using GIS
analysis techniques. Yet, can this conceptualizing of space reconcile with
Indigenous Science understandings? Although Western and Indigenous Science
may share some of these spatial thinking characteristics, the idea of a taxonomic
hierarchy of spatial thinking we regard as counter to the holistic sense of space
and place that is held by many Native Americans. We suggest that geoscientists
collaborate with Indigenous scientists to define spatial thinking in terms that are
context-sensitive. This approach might range from using local languages to re-
label the elements of Gersmehl’s taxonomy to de-constructing the notion of a
spatial hierarchy and replacing it with other context-specific definitions. These
could be derived from oral histories associated with particular locales, as was
the case in Semken’s report (2005) on an Indigenous Physical Geology course
and Palmer’s (2012) use of iGIN for creating a Kiowa GIS database. The notion of
a holistic sense of space must encapsulate a diversity of meanings to recognize
the degrees to which individuals find attachment to specific places (Semken &
Freeman, 2008). The boundaries of an area, the names given to geographic
features, and how they are interrelated and given value are among a myriad of
factors to consider when trying to define how individuals and groups may think
spatially about a particular locale. Palmer’s (2012) hybrid iGIN model blending
Indigenous and Western spatial knowledge can offer a way forward. To build on
this proposition, we offer the broader concept of spatial awareness as a descriptor

10
that may better suit the more nuanced, holistic understandings of space and how
it is linked to an individualized sense of place Although we acknowledge there
are likely to be potential dissimilarities between spatial thinking and spatial
awareness, we are cognizant of the risk of conflating these two concepts and
how that may be construed as deterministic; the Western view of spatial
thinking over-riding Indigenous Science’s sense of spatial awareness. We re-
purpose the concept of respatialization, defined as ―the transformation of spatially
referenced data from their original geographic representation to an alternative
geographic framework‖ (Goodchild & Janelle, 2010, p. 7) to characterize a
process where Native Americans frame their own, unique understanding of
what it means to think spatially in a fashion that is grounded in their local
context and language. This proposition is informed by the efforts of qualitative
geographers to challenge Geography’s positivist tradition (Harvey, 1984; Louis,
2007; Palmer, 2012; Pavlovksya, 2006).
Explicating the differences between these two perspectives and seeking
common ground is an important task for cognitive spatial researchers,
geoscientists and Indigenous Science practitioners and we recommend it as a
topic of further research. Nevertheless, common ground exists to build
consensus regarding what it means to think spatially. Western Science and
Native Science ontologies can be co-mingled, according to Cajete (1999), as
evidenced by a gradual recognition of indigenous knowledge by mainstream
science (Couzin, 2007; Semken & Freeman, 2008). Respect for indigenous ways
of knowing and a solid base of science knowledge and pedagogy should be
complementary (Semken & Freeman, 2008). Of the studies discussed in this
review, we regard Palmer’s (2012) iGIN model as an exemplar for future
research into how spatial knowledge can be conceptualized and labelled. It
argues for a nuanced hybridization of Western and Indigenous terminologies to
describe spaces and places and supports its claims with a description of an
indigenous-centric travel narrative map integrated into a conventional GIS.
7. Conclusion: Recommendations for Future Research
To foster and nurture STEM interest in Native American students, there
are a variety of approaches from geoscience research and practice that educators
may draw upon. GIS and geovisualization tools in place-based educational
program can not only promote interest in STEM education, they can be
congruent with Indigenous education practices, provided that both views of
what it means to think spatially are presented. As a final caveat, implementing
any STEM-focused, place-based educational program within rural Native
communities must be conducted with the express permission and contribution
of tribal members and governing bodies as equal partners in feasibility studies,
research, or implementation.
An awareness of the linkages between Indigenous Science and place-
based education allow these two practices to advance an alternative meaning of
space and place which is more localized. The important issue is how to reconcile
two competing scientific paradigms. We propose that spatial learning, as
exemplified by new technologies and research efforts in GIS and
geovisualization, offers innovative ways of examining and understanding what

11
it means to think spatially in locations that have relevance to students’
communities and daily lives. By utilizing the concept of respatialization the
terminology for spatial thinking espoused by Western Science, may be reclaimed
by Indigenous scientists. They can adopt names of places and geophysical
processes that have been passed down through oral traditions shaped by the
interaction of locale and cultural-linguistic traditions. If Native American
students contribute to this process by using GIS and geovisualization tools to
critically examine and catalogue their locales from a hybridized Western-
Indigenous Science spatial perspective, we believe that it can increase geoscience
and STEM interest. We also recommend that geoscientists, spatial-cognitive
scientists, and Indigenous scientists collaborate on research that recognizes the
variety of possible meanings and labels associated with thinking spatially.
Acknowledgements. This publication was made possible by the National
Science Foundation Idaho EPSCoR Program under award number IIA-1301792.
References
Adetunji, O. O., Ba, J.-C. M., Ghebreab, W., Joseph, J. F., Mayer, L. P. & Levine, R. (2012).
Geosciences awareness program: A program for broadening participation of
students in geosciences. Journal of Geoscience Education, 60(3), 234–240.
Aikenhead, G. S. (1998). Toward a First Nations cross-cultural science and technology
curriculum. Science Education, 81(2), 217-238. doi: 10.1002/(SICI)1098-
237X(199704)81:2<217::AID-SCE6>3.0.CO;2-I
Apple, J., Lemus, J., Semken, S. (2014). Teaching geoscience in the context of culture and
place. Journal of Geoscience Education, 62, 1-4.
Babco, E. L. (2003). Trends in African American and Native American Participation in STEM
Higher Education. Washington, DC: Commission on Professionals in Science and
Technology.
Baker, T. R., & White, S. H. (2003). The effects of GIS on students' attitudes, self-efficacy,
and achievement in middle school science classrooms. Journal of
Geography, 102(6), 243-254.
Bednarz, S. W. (2004). Geographic information systems : A tool to support geography
and environmental education ? GeoJournal, 60, 191–199.
Bednarz, R.S & Lee, J. (2011). The components of spatial thinking: Empirical evidence.
Procedia-Social and Behavioral Sciences, 21, 103-107.
doi:10.1016/j.sbspro.2011.07.048
Bolles, R. C., Baker, H. H., & Marimont, D. H. (1987). Epipolar-plane image analysis: An
approach to determining structure from motion. International Journal of Computer
Vision, 1(1), 7-55.
Burns, J. H. R., Delparte, D., Gates, R. D., & Takabayashi, M. (2015). Integrating
structure-from-motion photogrammetry with geospatial software as a novel
technique for quantifying 3D ecological characteristics of coral reefs. PeerJ, 3,
e1077.
Cajete, G. (1994). Look to the mountain: An ecology of indigenous education. Durango, CO:
Kivaki Press.

12
Cajete, G. (1999). The Native American learner and bicultural science education. Next
steps: research and practice to advance Indian education (pp. 135–160). Retrieved
from http://www.eric.ed.gov/ERICWebPortal/recordDetail?accno=ED427908
Cajete, G. (2000). Indigenous knowledge: The Pueblo metaphor of indigenous
education. Reclaiming Indigenous voice and vision, 181-191.
Castagno, A. E., & Brayboy, B. (2008). Culturally-responsive schooling for Indigenous
youth: A review of the literature. Review of Educational Research, 78(4), 941-993.
Chemers, M. M., Zurbriggen, E. L., Syed, M., Goza, B. K., & Bearman, S. (2011). The role
of efficacy and identity in science career commitment among underrepresented
minority students. Journal of Social Issues, 67(3), 469-491.
Couzin, J. (2007). Opening doors to native knowledge. Science, 315(5818), 1518-1519. doi:
10.1126/science.315.5818.1518.
DeFelice, A., Adams, J. D., Branco, B., & Pieroni. P. (2014) Engaging underrepresented
high school students in an urban environmental and geoscience place-based
curriculum. Journal of Geoscience Education: February 2014, Vol. 62, No. 1, pp. 49-
60. doi: http://dx.doi.org/10.5408/12-400.1
Deloria, V. (2003). God is Red: A Native View of Religion. Golden, CO: Fulcrum Publishing.
Delparte, D.M., Johnson, J. & Tracy, M. (in prep). Re-envisioning the Portneuf River
corridor with geodesign.
Delparte, D. M., Belt, M., Nishioka, C., Turner, N., Richardson, R. T., & Ericksen, T.
(2014). Monitoring tropical alpine lake levels in a culturally sensitive
environment utilizing 3D technological approaches. Arctic, Antarctic, and Alpine
Research, 46(4), 709-718.
Doering, A. & Veletsianos, G. (2007). An Investigation of the Use of Real-Time, Authentic
Geospatial Data in the K – 12 Classroom. National Council for Geographic
Education, 106, 217–225.
Downs, R., & DeSouza, A. (2006). Learning to think spatially: GIS as a support system in
the K-12 curriculum. Committee on the Support for the Thinking Spatially, National
Research Council, The National Academies Press, URL: http://books. nap. edu/catalog.
php.
Fleishman, E. A. C. & Rich, S. (1963). Role of kinesthetic and spatial-visual abilities in
perceptual-motor learning. Journal of Experimental Psychology, 66(1), 6-11.
Gersmehl, P.J. & Gersmehl, C. A. (2006). Wanted: A concise list of neurologically
defensible and assessable spatial thinking skills. Research in Geographic Education,
8(1), 5-38.
Goodchild, M. F. & Janelle, D. G. (2010). Toward critical spatial thinking in the social
sciences and humanities. GeoJournal, 75(1), 3–13.
Grossman, J. M., & Porche, M. V. (2013). Perceived gender and racial/ethnic barriers to
STEM success. Urban Education, 0042085913481364.
Harvey, D. (1984). On the history and present condition of geography: an historical
materialist manifesto. The Professional Geographer, 36(1), 1-11.
Hauptman, H., & Cohen, A. (2011). The synergetic effect of learning styles on the
interaction between virtual environments and the enhancement of spatial
thinking. Computers & Education, 57(3), 2106–2117.
Johnson, A. N., Sievert, R., Durglo Sr, M., Finley, V., Adams, L. & Hofmann, M. H.
(2014). Indigenous Knowledge and Geoscience on the Flathead Indian
Reservation, Northwest Montana: Implications for Place-Based and Culturally
Congruent Education. Journal of Geoscience Education, 62(2), 187–202.

13
Kaltenborn, B. P. (1998). Effects of sense of place on responses to environmental impacts:
A study among residents in Svalbard in the Norwegian high arctic. Applied
Geography, 18(2), 169-189.
Koenderink, J. J., & Van Doorn, A. J. (1991). Affine structure from motion. JOSA A, 8(2),
377-385.
Lee, J., & Bednarz, R. (2009). Effect of GIS learning on spatial thinking. Journal of
Geography in Higher Education, 33(2), 183-198.
Lieberman, G. A., Hoody, L. L. (1998). Closing the Achievement Gap: Using the Environment
as an Integrating Context for Learning. San Diego, CA: State Education and
Environmental Roundtable.
Louis, R. P. (2007). Can you hear us now? Voices from the margin: Using indigenous
methodologies in geographic research. Geographical research, 45(2), 130-139.
Lynch, K., Bednarz, B., Boxall, J., Chalmers, L., France, D., & Kesby, J. (2008). E-learning
for geography's teaching and learning spaces. Journal of Geography in Higher
Education, 32(1), 135-149.
Palmer, M. (2012). Theorizing indigital geographic information networks. Cartographica:
The International Journal for Geographic Information and Geovisualization, 47(2), 80–
91.
Pavlovskaya, M. (2006). Theorizing with GIS: a tool for critical geographies? Environment
and Planning A, 38(11), 2003.
Pewewardy, C. (2002). Learning styles of American Indian/Alaska Native students: A
review of the literature and implications for practice. Journal of American Indian
Education, 41(3), 22-56.
Richardson, T. R. (in prep). User performance and cognitive load measures of
geovisualization and tactile feedback maps: A comparative analysis. Ph.D.
dissertation, Idaho State University.
Riggs, E. M., Robbins, E. & Darner, R. (2007). Sharing the land: Attracting Native
American students to the Geosciences. Journal of Geoscience Education, 55(6), 478-
485.
Riggs, E. M., & Semken, S. C. (2001). Earth Science for Native Americans-An endogenic-
exogenic process is also an Earth-Sky relationship, if you're learning geoscience
at the Dine Tribal College in New Mexico. Geotimes, 46(9), 14-17.
Morgan, F., & Semken, S. C. (1997). Navajo pedagogy and Earth systems .Journal of
Geoscience Education, 45, 109.
Semken, S. (2005). Sense of place and place-based introductory geoscience teaching for
American Indian and Alaska Native undergraduates. Journal of Geoscience
Education, 53(2), 149-157.
Semken, S., & Freeman, C. B. (2008). Sense of place in the practice and assessment of
place-based science teaching. Science Education, 92(6), 1042–1057.
doi:10.1002/sce.20279
Semken, S., Freeman, C. B., Watts, N. B., Neakrase, J. J., Dial, R. E., & Baker, D. R. (2009).
Factors that influence sense of place as a learning outcome and assessment
measure of place-based geoscience teaching. Electronic Journal of Science
Education, 13(2), 136–159.
Shamai, S. (1991). Sense of place: An empirical measurement. Geoforum, 22(3), 347-358.

14
Schultz, R. B., Kerski, J. J., & Patterson, T. C. (2008). The use of virtual globes as a spatial
teaching tool with suggestions for metadata standards. Journal of Geography, 107,
27–34.
Snively, G., & Corsiglia, J. (2001). Discovering indigenous science: Implications for
science education. Science Education, 85(1), 6-34.
Sobel, D. (2004). Place-based education: Connecting classroom and community. Nature
and Listening, 4. Retrieved from http://www.antiochne.edu/wp-
content/uploads/2012/08/pbexcerpt.pdf
Syed, M., Goza, B. K., Chemers, M. M., & Zurbriggen, E. L. (2012). Individual
differences in preferences for matched-ethnic mentors among high-achieving
ethnically diverse adolescents in STEM. Child development, 83(3), 896–910.
doi:10.1111/j.1467-8624.2012.01744.x
Titus, S., & Horsman, E. (2009). Characterizing and improving spatial visualization skills.
Journal of Geoscience Education, 57(4), 242–254.
Williams, D. R., & Vaske, J. J. (2003). The measurement of place attachment: Validity and
generalizability of a psychometric approach. Forest science, 49(6), 830-840.
Velez, M. C., Silver, D. & Tremaine, M. (2005). Understanding visualization through
spatial ability differences. In Visualization, 2005. VIS 05. IEEE, 511–518.
Vogler, R., Ahamer, G., & Jekel, T. (2010). GEOKOM-PEP. Pupil led research into the
effects of geovisualization. Learning with GI, V, 51–60.
Wang, X. (2013). Why students choose STEM majors: Motivation, high school learning
and postsecondary context of support. American Educational Research Journal,
50 (1081). doi: 10.3102/0002831213488622. Retrieved from
https://www.insidehighered.com/sites/default/server_files/files/Wang%20A
ERJ%20Oct%202013.pdf
Zalles, D. R., Collins, B. D., Montgomery, D., Colonesese, T., & Updegrave, C. (2005).
Merging Place-Based environmental science and traditional ecological knowledge in
secondary and postsecondary educational settings. Retrieved from
http://digers.sri.com/downloads/DIGERS_Paper_National_Association_for_R
esearch_in_Science_Teaching_March2012.pdf

16
© 2016 The authors and IJLTER.ORG. All rights reserved.
Vol. 15, No. 5, pp. 16-26, April 2016
Using Coh-Metrix to Analyze Chinese ESL
Learners’ Writing
Weiwei Xu
College of International Studies,
Southwest University,
Chongqing, China
Ming Liu
School of Computer and Information Science,
Southwest University,
Chongqing, China
Abstract. Scoring essays is costly, laborious and time-consuming.
Automated scoring of essays is a promising approach to face this
challenge. Coh-Metrix is a computer tool that reports on cohesion,
sentence complexity, lexical sophistication and other descriptive
features at sentence- and paragraph-level. It has been widely used to
analyze native English speakers’ essay writing. However, few studies
have used Coh-Metrix to analyze essays written by English as a Second
Language (ESL) students. In this study, we analyzed the correlation
between several Coh-Metrix features combined with a set of newly
proposed features and the quality of essays, written by Chinese
university students, both English and non-English majors. This study
shows that each group of students tends to write essays that have their
own signature features. The quality of essays written by English majors
highly correlate to the importance of introduction, conclusion and
cohesion at the sentence level, while the quality of essays written by
chemistry majors are highly related to mechanics errors, sentence
complexity and cohesion at the paragraph level.
Keywords: ESL essay writing, Textual feature analysis, Automatic Essay
Scoring, Computer in education
Introduction
Important constructs, central to ESL writing and proposed by several
researchers, are grammatical and spelling errors. Cohesion is also important,
although it is a much more difficult aspect of writing to account for due to its
deeper nature (Rus & Niraula, 2012). This study focuses on grammatical and
spelling errors and cohesion which are directly observed through the explicit
presence or absence of specific tokens. Errors may be caused by inappropriate
transfer of first language patterns and/or incomplete knowledge of the target

17
language, in this case, English. Researchers (Q. Liang, 2004; Liu, 2008) have
pointed out that Chinese college students, especially those with low proficiency
in English, often make errors at the surface level, such as spelling and
grammatical errors (e. g. run-on sentences); errors at high level, such as using
Chinglish (ungrammatical English expressions used in Chinese context, having
deprecating connotation); and low cohesion or incorrect use of connectives. Even
for students with high proficiency, like English majors, writing high quality
essays with high cohesion, well-established introduction and conclusion,
remains a challenge. Thus, a marking tool, specifically developed to analyze ESL
learners’ errors, is very much needed. It should be noted that errors are
categorized as word-level (spelling errors) and sentence-level (grammatical
errors) and, as mentioned above, are consequences of incomplete knowledge of
the target language or of the transfer difficulty due to major dissimilarities
between the foreign language and students’ native language. On the other hand,
cohesion is a discourse-level aspect of writing and lack of cohesion in an essay
may reflect lack of composition training and practice. This distinction is
important to make, because one can argue that the only net advantage of native
speakers of English over ESL speakers is their knowledge of English vocabulary
and grammar. Discourse-level aspects, on the other hand, are governed by
general cross-language principles of cohesion and coherence, and are equally
impacting for both native and EFL speakers. As is shown in this study, English
majors who presumably have mastered the mechanics of the language
(vocabulary and grammar) struggle mainly with the compositional aspect,
which is in contrast with non-English majors who struggle with both the
mechanics and composition aspect of essay writing.
Researches in computer-based essay scoring, referred to as Automatic Essay
Scoring (AES), have been going on for more than 40 years. The first known AES
system, called Project Essay Grader (Page, 2003) based on a regression model,
was developed by Ellis Page in 1966. With the advancement of Natural
Language Processing (NLP) and Information Retrieval (IR) techniques, four
more advanced AES systems were developed during the late 1990s (M. Shermis
& Burstein, 2003). In recent years, different approaches to AES were proposed
(McNamara, Crossley, Roscoe, Allen, & Dai, 2015; Mark D. Shermis, 2014). AES
systems in China is still at an early stage (Ge & Chen, 2007; Han, 2009; Li, 2009;
M. Liang & Wen, 2007; M. Liang, 2011). Most of researchers focus on the reviews
of existing AES systems and their potential applications to Chinese ESL context
(Ge & Chen, 2007; Han, 2009; M. Liang & Wen, 2007). Few researchers (Li, 2009;
M. Liang, 2011) have attempted to develop AES systems in Chinese ESL context
by using latent semantic analysis technique (Landauer, Foltz, & Laham, 1998).
This paper aims to explore what textual features are good predictors for writing
quality and investigate its implication for developing AES system in Chinese
ESL context. Textual features such as syntactic patterns, cohesions and
connectives were extracted by using the computational tool Coh-Metrix
(Graesser, McNamara, Louwerse, & Cai, 2004). Coh-Metrix is used to analyze
essays written by Chinese ESL students, and this study analyzed the correlations
between features and the quality of essays written by both English and non-
English majors.

18
Coh-Mextrix
Coh-Metrix is a computational tool that provides over 100 indices of cohesion,
syntactical complexity, connectives and other descriptive information about
content (Graesser et al., 2004). Due to space restriction, only a summary of Coh-
Metrix’s key features is presented here. The current public version available is
Coh-Metrix 3.0, which can retrieve 108 scores of textual features. More
information can be found on the website
http://cohmetrix.Memphisedu/cohmetrixpr/index.html. A wide-range
overview is provided in (Graesser et al., 2004):
Descriptive Indices: It includes the number of paragraphs, sentences, words,
syllables in words, etc.
Cohesion: It is a key aspect for understanding the discourse structure of a
language and how connectives used in a text have an impact on cohesion
(Kintsch & van Dijk, 1978).
Sentence Complexity: It indicates human graders’ evaluations of the quality of the
text.
Lexical Sophistication: It refers to the writer’s use of advanced vocabulary and
word choice to express his or her thought.
New Features
This study proposes and extracts 8 new features that are not available in Coh-
Metrix. These features refer to characteristics of ESL learners’ writing styles and
reflect on the importance of the introduction section, conclusion section and
mechanics in errors including spelling and grammatical errors. Students often
make the mistake of jumping straight to answering the essay question in the first
paragraph without following a background statement, essay statement or
outline statement. In addition, students rush to finish up in conclusion. The
conclusion section should restate the author's stance with respect to the essay
question, make a brief summary of evidences and finish with some sort of
judgment about the topic. Moreover, spelling and grammatical errors are always
good indicators of essay quality.
Number of Words in Introduction: the total number of words in the first paragraph
considered as introduction.
Number of Words in Conclusion: the total number of words in the last paragraph
considered as conclusion.
Introduction Portion: the ratios of number of words in introduction to the total
number of words in the essay.
Conclusion Portion: the ratios of number of words in the conclusion to the total
Spelling Errors: the number of spelling errors. This study employs an open source
spelling error checker called Language Tool (http://www.languagetool.org/),
which is a part of the Open Office suite.
Grammatical Errors: the number of sentences with grammatical errors. This study
uses the Link Grammar Parser (Lafferty, Sleator, & Temperley, 1992) to check
the grammar of a sentence, which is also widely used in ESL context.
Percentage of Spelling Errors: the ratios of the number of spelling errors to the total

19
Percentage of Grammatical Errors: the ratios of the number of sentences with
grammatical errors to the total number of sentences in the essay.
Methodology
Participants
Essays were collected from 90 freshmen at one of China’s key universities.
Among them, 41 students were English majors at the College of International
Studies, while 49 students were chemistry majors at the School of Chemistry.
English majors are considered to have the higher English proficiency. For the
English majors, their average score in English as a testing subject in the National
Higher Education Entrance Examination (also called Gaokao) was 131.30, and
the standard deviation was 7.37. For the chemistry majors, their average score
was 110, and the standard deviation was 10.14. Three experienced English
teachers at the College of International Studies at the university volunteered to
rate the quality of essays. All of them have at least five years of experience in
teaching a composition course for both English and non-English majors.
Task and Instruments
The writing task was timed and considered as an assignment in English class.
Students were required to finish it within 30 minutes. The writing task was to
write a persuasive essay following the standard of college English essay writing
set by the Ministry of Education in China.
The essays were rated by the three experienced English teachers mentioned
above. They evaluated students’ essays based on the standardized rubric
commonly used to grade college English essay tests on the scale of 1 to 100. They
first evaluated 18 essays. If the correlations between the teachers did not exceed
r=.50 on each item, the evaluation process were rechecked until correlation was
greater an equal to 0.5. After they reached a moderate agreement, each teacher
then evaluated the 72 essays that comprised the whole sample used in this
study.
It was found out that their inter-rater reliability was high with r=.756, r=.745,
r=.607, respectively, p<.001. The scoring rubric included organization (e. g. clear
organization of subtopics), content (e. g. clearly expressing ideas, text coherence,
interesting and balanced introduction and conclusion) and mechanics (e. g.
errors in punctuation and grammar).
These essays were chosen because the types they represented better reflected the
conditions under which students usually completed prompt-based essays, such
as CET or TEM. In addition, these two student groups can be representatives of
most of the university students including English majors and non-English
majors. Hence, the results of the selected features and algorithms are more likely
to be accurate in the context of Chinese ESL writing. Indeed, the English majors’
essays exhibit more discourse-level issues, while the non-English majors’ essays
exhibit both basic-level issues (spelling- and grammar-level) and discourse-level
issues. This is the case due to English majors’ more knowledge about the basics
of the target language, English.
Results and Discussion
Descriptive statistics for the English majors and chemistry majors as well as the
hybrid group (the combination of both essays) are reported in TABLE I.

20
The average scores of the English majors’ and chemistry majors’ essays were not
significantly different. The English majors and chemistry majors’ essays were
significantly different when number of words, sentences, paragraphs and
syllables per word are involved. It indicates that the essays written by the
English-major students contain more words, sentences and paragraphs, but less
complicated words (less syllables), compared with the essays written by the
chemistry majors. In addition, the English majors made less grammatical and
spelling errors than the chemistry majors did.
Key Features for English Majors’ Essays
Top six features were selected by using the same feature selection method used
above, but this time applied on the training set (21 essays) written by the English
majors. The linear regression yielded a significant model, F(6,14)=10.982,
p<0.001, r=.944, r2=.892. Table II shows the six features that correlate with the
essay scores. The conclusion portion was positively related to essay quality. But,
the feature of the introduction portion was negatively related to the scores. It
indicates the importance of the summarization of arguments in the final section
of essays, as found in previous study (Freeman & Freeman, 1998). Cohesion as
measured by content word overlap and Wordnet overlap were positively related
to essay quality, which was similar to the results reported in a previous study
(SA Crossley & McNamara, 2010). However, the argument overlap was
TABLE I. DESCRIPTIVE AND ANOVA STATISTICS FOR ENGLISH MAJORS’ AND
CHEMISTRY MAJORS’ ESSAYS IN THE DATASET
Features
English
Majors
Chemistry
Majors
F(1,71) Hybrid
Raters’ Essay
Evaluations
70.45(9.95) 73.30(7.64) 1.362 72.10(8.72)
Number of Words 274(46.28) 136.47(4.00) 203.95* 194.65(76.56)
Number of
Sentences
17.23(3.87) 9.30(0.52) 72.23* 12.65(5.14)
Number of
Paragraphs
4.41(0.79) 3.03(1.03) 27.12* 3.62(1.16)
Number of Syllables
per word
1.41(0.06) 1.61(0.10) 73.03* 1.53(1.31)
Number of Spelling
Errors per
Document of Words
2.9(1.94) 3(2.91) 8.77* 2.94(2.37)
Number of
Grammar Errors per
Document of
Sentences
4.23(2.28) 6(2.23) 6.05* 4.98(2.40)

21
negatively correlated to essay quality. Argument overlap occurred when there
were matching personal pronouns between sentences. It is observed that
unskillful writers like to use a person’s experience as an example to support the
arguments in an illogical way. These essays contain many pronouns such as
“he” and “his”. The following text segment is extracted from one poor quality
essay from the dataset. Although this example essay has high argument overlap,
it lacks logic between the following two adjacent sentences: “My friend Bob, he
often helped his parents do household jobs and got reward when he was young.
So up to now, he always the best person I think, his experience makes him learn
how to independent.”
“Temporal Connectives” was negatively related to essay quality, because some
poor-skilled writers incorrectly used some temporal connectives, such as
“when”, “since” and “as”. As expected, English majors’ essays showed issues
at the discourse level such as temporal connectives and argument overlap which
negatively correlated with the quality of the essays.
Regression Model Performance in the English Major Group
In order to validate the regression model consisting of six features, the model in
these test sets (11 essays) were evaluated. It yielded r=.784, r2=.615. Therefore,
this result demonstrates that the combination of six features account for 61.5% of
the variance in the test set.
Categorical scores, including “distinction” (80-100), “credit” (70-79), “pass” (60-
69) and “fail” (0-59), are also one of the common credit systems used at China’s
universities, such as Southwest University (University, 2007). These categorical
scores are also used in many writing tests (Lawrence M. Rudner & Liang, 2002).
The scores derived from the test set were used to assess categorical accuracy of
the regression scores, compared with the human-graded scores. The regression
model produced categorical matches for 7 of the 11 essays (64 % accuracy). The
TABLE II. CORRELATIONS BETWEEN FEATURES AND RATERS’ SCORES IN
ENGLISH MAJOR GROUP IN THE TRAINING SET
Feature Type R P
Introduction Portion New feature -.635 <.050
Conclusion Portion New feature .576 <.050
Argument Overlap Cohesion -.551 <.050
Content Word Overlap Cohesion .521 <.050
Temporal Connectives Cohesion -.803 <.001
WordNet Overlap between
Verbs
Cohesion .714 <.050

22
reported, weighted Cohen’s kappa for the categorical matches was 0.516,
demonstrating a moderate agreement. A confusion matrix for this analysis is
provided in TABLE III.
Key Features for the Chemistry Majors’ Essays
The top seven features were selected by using the same feature selection
algorithm as before, but this time applied on the training set (27 essays) written
by the students majoring in chemistry. The linear regression yielded a significant
model, F(7,19)=3.186, P <0.05, r=.709, r2=.503. TABLE IV presents the
correlations between these features and scores. Among these features, it is
observed that results are similar to those reported in other studies (Scott a.
Crossley & McNamara, 2011; McNamara, Crossley, & Roscoe, 2013; Mcnamara,
Crossley, & Mccarthy, 2010). Essay quality is positively related with essay length
(number of words) and cohesion (semantic similarity between adjacent
paragraphs). As expected, the new features “Percentage of Spelling Errors” and
“Percentage of Grammatical Errors” were negatively related to the essay quality.
Surprisingly, the syntactic complexity (incidence score of verbal phrases) was
negatively related to essay quality, which was different to the results found in
the previous study (Mcnamara et al., 2010). This may be one characteristic of
ESL writers, since they are more likely to make grammatical mistakes if they try
to write complex sentences. Another cohesion feature, “Standard Deviation of the
Semantic Similarity between Sentences”, showed negative correlations with essay
quality. It indicates that the semantic inconsistency between sentences was
negatively correlated to essay quality. Unlike studies in the past, there is a
negative correlation between “Logical Connectivity” and essay scores. It is found
out that many essays with poor marks had many “and” as a logical connective.
It was used almost always for connecting two nouns or adjectives, such as “more
and more popular”, and “China and the West”. As expected, the non-English
majors show more problems at basic-levels of writing, such as spelling and
grammatical errors.
TABLE III. HUMAN CATEGORICAL SCORE PREDICTION IN THE ENGLISH MAJOR
GROUP IN THE TEST SET
System Predicted
Scores
Actual Human Scores
Distinction Credit Pass Fail
Distinction 2 0 0 0
Credit 0 2 0 0
Pass 0 1 1 1
Fail 0 1 1 2

23
Regression Model Performance in the Chemistry Major Group
In order to validate the regression model consisting of seven features, this model
in the test set (13 essays) written by the chemistry majors was evaluated. It
yielded r=.569. The scores derived from the test set were used to assess the
categorical accuracy of the regression scores, compared with the human-graded
scores. The regression model produced categorical matches for 8 of the 13 essays
(54 % accuracy). The reported, weighted Cohen’s kappa for the categorical
matches was 0.404, demonstrating a moderate agreement. A confusion matrix
for this analysis is provided in TABLE V.
This matrix reflects a decrease in the categorical agreement using the model
tested in the dataset of the chemistry majors’ essays. The predicted scores tend to
be in the “credit” category since 8 of the 13 essays have been predicted in the
“credit” category. This level of performance is partially due to the frequent
credit scores and small variations of actual human scores (SD: 7.64), which
renders the prediction task more difficult.
TABLE IV. CORRELATION OF THE FEATURES AND RATERS’ SCORES IN THE
CHEMISTRY MAJOR GROUP IN THE TRAINING SET
Feature Type R P
Number of Words Descriptive .676 <.050
Percentage of Spelling Errors New Feature -.486 <.050
Percentage of Grammatical
Errors
New Feature -.460 <.050
Logical Connectivity Cohesion -.450 <.050
Standard Deviation of the
Semantic Similarity between
Sentences
Cohesion -.531 <.050
Semantic Similarity between
Adjacent Paragraphs
Cohesion +.528 <.050
Incidence Score of Verbal
Phrases
Syntactic
Pattern
-.641 <.050
TABLE V. HUMAN CATEGORICAL SCORE PREDICTION IN THE
CHEMISTRY MAJOR GROUP IN THE TEST SET
System
Predicted
Scores
Actual Human Scores
Distinction Credit Pass Fail
Distinction 2 0 0 0
Credit 0 2 0 0
Pass 0 1 1 1
Fail 0 1 1 2

24
Conclusion and Future work
This study has used a set of Coh-Metrix indices combined with a set of newly
proposed features to analyze ESL essays written by the English majors and non-
English majors at a university in China. It showed the predictive values of
several features extracted using Coh-Metrix; some of the newly proposed
features significantly correlated to essay quality as well. These features include
cohesion at the sentence- and paragraph-level, introduction and conclusion
portion, syntactical complexity and surface errors. The results indicate the
usefulness of Coh-Metrix and the newly proposed new features. Interestingly,
different features are more significant for different groups of essays. The English
majors emphasize cohesion between sentences, writing a good summarization,
whereas the non-English majors focus on making less surface errors, such as
spelling and grammatical errors, and cohesion between adjacent paragraphs.
This study has some limitations. For example, the sample size is not big enough,
since 72 essays and two groups of ESL writers were analyzed. However, these
essays were written by university students in a real scenario and the data
analysis process was sound. In the future, improving the performance of the
prediction model will be the focus. At the present, most of the studies use a
linear regression model for essay score prediction. Non-linear regression
models, such as SVM Regression (Shevade, Keerthi, Bhattacharyya, & Murthy,
1999) and other machine learning techniques (Hongbo Chen, Ben He, Tiejian
Luo, 2012; Larkey, 1998) will be investigated. Moreover, more ESL essays
written by university students from different disciplines will be collected and
analyzed.
Acknowledgement:
This article was supported by Chongqing Social Science Planning Fund Program
[2014BS123], Fundamental Research Funds for the Central Universities
[XDJK2014A002], [XDJK2014C141] and [SWU114005] in China.
References
Crossley, S. A., & McNamara, D. S. (2011). Understanding expert ratings of essay quality:
Coh-Metrix analyses of first and second language writing. International Journal of
Continuing Engineering Education and Life-Long Learning, 21(2/3), 170.
http://doi.org/10.1504/IJCEELL.2011.040197
Crossley, S., & McNamara, D. (2010). Cohesion, coherence, and expert evaluations of
writing proficiency. In The 32nd Annual Conference of the Cognitive Science Society
(pp. 984–989). Austin: TX. Retrieved from
http://csjarchive.cogsci.rpi.edu/Proceedings/2010/papers/0310/paper0310.pdf
Freeman, Y. S., & Freeman, D. E. (1998). ESL/EFL Teaching: Principles for Success.
Heinemann.
Ge, S., & Chen, X. (2007). Automated Essay Scoring for Chinese EFL learners. Foreign
Language World, 122(5), 43–50.
Graesser, A. C., McNamara, D. S., Louwerse, M. M., & Cai, Z. (2004). Coh-metrix:
analysis of text on cohesion and language. Behavior Research Methods, Instruments, &
Computers, 36(2), 193–202. Retrieved from
http://www.ncbi.nlm.nih.gov/pubmed/15354684

25
Han, N. (2009). The theory and practice of Automated English Essay Scoring Systems.
China Test, 3, 38–44.
Hongbo Chen, Ben He, Tiejian Luo, B. L. (2012). A ranked-based learning approach to
automated essay scoring. In 2012 Second International Conference on Cloud and Green
Computing (CGC) (pp. 448–455).
Kintsch, W., & van Dijk, T. (1978). Towards a model of text comprehension and
production. Psychological Review, 85, 363–394.
Lafferty, J., Sleator, D., & Temperley, D. (1992). Grammatical Trigrams: A Probabilistic
Model of Link Grammar. In Proceedings of the AAAI Conference on Probabilistic
Approaches to Natural Language.
Landauer, T. K., Foltz, P. W., & Laham, D. (1998). Introduction to latent semantic
analysis. Discourse Processes, 25, 259–284.
Larkey, L. S. (1998). Automatic essay grading using text categorization techniques.
Proceedings of the 21st Annual International ACM SIGIR Conference on Research and
Development in Information Retrieval - SIGIR ’98, (1), 90–95.
http://doi.org/10.1145/290941.290965
Lawrence M. Rudner, & Liang, T. (2002). Automated essay scoring using Bayes’
Theorem. The Journal of Technology, Learning and Assessment, 1(2).
Li, J. (2009). Using latent semantic analysis for automated essay scoring in the Chinese EFL
context. Guangdong University of Foreign Studies.
Liang, M. (2011). Construting a model for the computer assisted scoring of Chinese EFL
learners’ argumentative essays. Foreigh Language Teaching and Research Press.
Liang, M., & Wen, Q. (2007). A critical review and implications of some automated essay
scoring systems. Computer Assisted Foreign Language Education, 18-24.
Liang, Q. (2004). Contrastive study on the objectivity of English and Chinese
argumentative writing: A survey on the objectivity of Chinese college students’
English argumentative writing. Journal of China West Normal University (Philosophy
& Social Sciences), (5).
Liu, X. (2008). A case study of a non-English Major’s Writing: Error analysis. Foreign
Language Research, 2, 140–142.
McNamara, D. S., Crossley, S. a, & Roscoe, R. (2013). Natural language processing in an
intelligent writing strategy tutoring system. Behavior Research Methods, 45(2), 499–
515. http://doi.org/10.3758/s13428-012-0258-1
Mcnamara, D. S., Crossley, S. A., & Mccarthy, P. M. (2010). Linguistic features of writing
quality. Written Communication, 27(1). http://doi.org/10.1177/0741088309351547
McNamara, D. S., Crossley, S. A., Roscoe, R. D., Allen, L. K., & Dai, J. (2015). A
hierarchical classification approach to automated essay scoring. Assessing Writing,
23, 35–59. http://doi.org/10.1016/j.asw.2014.09.002
Page, E. B. (2003). Project essay grade: PEG. In M. D. Shermis & J. Burstein (Eds.),
Automated essay scoring: A cross-disciplinary perspective (pp. 43–54). Mahwah, NJ:
Lawrence Erlbaum Associates.
Rus, V., & Niraula, N. (2012). Automated detection of local coherence in short
argumentative essays based on centering theory. In 13th International Conference on
Computational Linguistics and Intelligent Text Processing (pp. 450–461). New Delhi,
India.

26
Shermis, M., & Burstein, J. (2003). Automated essay scoring: A cross-disciplinary perspective.
Mahwah NJ: Lawrence Erlbaum Associates.
Shermis, M. D. (2014). State-of-the-art automated essay scoring: Competition, results,
and future directions from a United States demonstration. Assessing Writing, 20, 53–
76. http://doi.org/10.1016/j.asw.2013.04.001
Shevade, S., Keerthi, S., Bhattacharyya, C., & Murthy, K. (1999). Improvements to the
SMO Algorithm for SVM Regression. IEEE Transactions on Neural Networks.
University, S. (2007). Implementation guidances on the reform of credit system in Southwest
University. Chongqing. Retrieved from agronomy.swu.edu.cn/up/87.doc

27
© 2016 The author and IJLTER.ORG. All rights reserved.
Vol. 15, No. 5, pp. 27-42, April 2016
The Factors Affecting the Adaptation of Junior
High School Students with Severe Disabilities to
Inclusive or Segregated Educational Settings
Li Ju Chen
Chang Gung University, Taiwan
Abstract. The aim of this research is to explore the factors of
the adaptation of junior high school students with severe
disabilities (SD) to inclusive or segregated school
environments. The study was based on survey data gathered
from 868 students with SD who were studying in junior high
schools of Taiwan. The research found that: (1) Language,
cognitive, and visual abilities are key factors for succeeding
in an inclusive education setting; (2) Language skills are
correlated with successful adaptation for students with SD;
(3) Children with certain types of disabilities are diagnosed
later than children with other disabilities and therefore
receive intervention later; (4) The relationships among
intervention timing, language skills, and school adaptation
for children with SD vary by disability types. There are
implications for improving interventions for SD based on
these research findings.
Keywords: early intervention, inclusive school, intervention
timing, language skills, students with severe disabilities
Introduction
The National Center for Educational Statistics reported that fifty
percentage of students with disabilities spent more than eighty
percentage of their time in the general education system (Madden,
2012; Michael & Trezek, 2006). Educators believe social skills are
crucial to effectively integrate students with disabilities into the
general education system. It showed that children with disabilities
who study in typical life circle have developed more positive social
behaviors than the children studying in more segregated contexts
(Alquraini & Gut, 2012; Koegel, Koegel, Frea, & Fredeen, 2001).
Moreover, inclusive education allows students with disabilities to
interact with typical students, which prevents students from being
labeled. In past decades, many countries, including the United States
and Taiwan, enacted some regulations to ensure that students with

28
disabilities are included in typical education system (Alquraini &
Gut, 2012; Special Education Act of Taiwan, 2009).
The exterior placement of students with disabilities into general
classrooms does not mean a meaningful inclusion (Brown,
Ouellette-Kuntz, Lysaght, & Burge, 2011). Educational adaptation
thus is an important clue for evaluating whether the education
system is proper for the children or not. It is suggested to develop a
way to assess a child’s adaption in school.
Some researches argued that inclusive education should be
insisted only when the children could achieve positive academic
(Oliver, 2008; Rous, Hallam, McCormick, & Cox, 2010). If a child
cannot adapt well to a mainstream environment, transferring to a
more segregated learning setting might be a more appropriate
placement for the student to have improved academic experience.
The disability of a child should be considered to lead him into certain
activities (World Health Organization, 2001). If an education system
can afford the students an environment to take part in more school
activities than another environment, then the first one is more
appropriate for the student than the second. The student is perceived
as having a milder disability in the first one than in the second
setting. Koegel et al. (2001) and Huang (2003) studied the effect of
school adaptation on student’s interactions with their classmates and
teachers and on participating in activities. It is claimed that these
themes are important supports by the schools when the students
study in inclusive schools (Kurth et al., 2015). Based on the literature
review, this research evaluate students’ overall school adaptation by
their academic progress, activity participation, and social
relationships.
Intervention Timing and School Adaptation. This study
explored the factors that promoted a student’s adaptation to school
in an inclusive or segregated education system. It is thought that
early intervention facilitates the children with disabilities adapt to
inclusive school (Low & Lee, 2011). Many studies have demonstrated
that intensive preschool intervention brings various benefits,
including academic, social, and economic issues, and enables the
children adapt to inclusive education setting (Zucker, 2010).
Intervention during children’s infancy and preschool stages has
produced aggressive results and promote the children’s educability
(Rogow, 2005). Neuman (2007) concluded that interventions are more
effective the earlier they are made. Several studies have indicated
that identification and intervention in time can avoid development
problems and promote developmental outcomes (Aron & Loprest,
2012; Puig, 2010; Renshaw et al., 2009; Shonkoff & Meisels, 2002).
Other studies have tried to identify the ideal intervention timing that
will maximize children adapting and learning well in an inclusive
education setting (Akshoomoff, Stahmer, Corsello, & Mahrer, 2013;
Stahmer, Carter, Baker, & Miwa, 2003). In accordance with previous

29
researches, the present research will explore the relationship
between intervention timing and subsequent school adaptation.
Language Skills and School Adaptation. Akshoomoff et al.
(2013) indicated that a child with a disability’s school adaptation
relates to communication skills. The child needs communication to
interact with others or participate in activities. An example is that
hallway greetings enable children to interact and initiate
conversations with other persons. These greetings require oral
language delivered (Rossetti, 2011). Language communication is
important in mainstream setting for conveying a variety of messages;
therefore, children should have ongoing opportunities to improve
language skills (Low & Lee, 2011; Puig, 2010; Rogow, 2005). The
1960s’ Head Start Program emphasized improving children’s
language ability to prevent them from learning failure in future
schooling. It is believed that students adapt better when they have
better language skills.
Intervention Timing and Language Skills. Interacting and
developing relationships with others in various contexts contribute
to a child’s language skills; children’s brain and their innate capacity
to develop language skills are stimulated by the persons interacting
with them (Puig, 2010). Many hospitals have established
intervention programs to provide additional stimulation and
organized activities for children with disabilities (Zucker, 2010).
Studies have shown that intervention timing and the acquisition of
language skills are related. Research in Norway, for instance, found
that eight-year-old children with disabilities who were involved
early intervention had better language ability than those who were
not involved. Akshoomoff et al. (2013) found that the children who
received early intervention obtained better scores in the
communication subscale of the Vineland Adaptive Behavior Scale.
According to the reviewed literature, the relationships among
early interventions, language skills, and school adaptation are
significant. Therefore, this research will also explore how
intervention timing, language skills and school adaptation are
related with one another among the students with disabilities.
Children with Disabilities in Different Types and Levels.
Many studies have claimed that intervention effectiveness,
intervention timing, language skills, and school adaptation vary
greatly with disability level and type. Neuman (2007) indicated that
interventions for children with mild disabilities are generally more
effective in intervention than for children with severer disabilities.
For example, the abilities required of students with mild visual
impairments (VI) to adapt to inclusive schools may be different from
those required of students with severe VI. Livneh and Wilson (2003)
found that life adaptation was impacted by disability level.
Statistical analyses examining all disability levels simultaneously
might lead to incorrect conclusions, the analyses of intervention
issues should be performed for various disability level individually.

30
Alquraini and Gut (2012) noted that greater part of studies have
focused on students with mild disabilities and advocated that more
topics be conducted with the students with severe disabilities (SD).
Some researches claimed to explore the critical components to
include the students with severe disabilities into typical educational
settings (Brock, Biggs, Carter, Cattey, & Raley, 2016; Kurth, Lyon, &
Shogren, 2015). The present study focuses on students with SD.
Inclusion setting afforded conditions for the students with SD to
develop relationships and social abilities by contacting with their
typical classmates (Alquraini & Gut, 2012). It is advocated to find the
practice factors supporting the students with SD to effectively study
in inclusive education setting (Brock, Biggs, Carter, Cattey, & Raley,
2016; Kurth, Lyon, & Shogren, 2015).
Children with disabilities in different types go through different
difficulties to school and social adaptation. Children with severe
cognitive impairment are worse at language of reception and
expression (Alberta Education, 2009). Most of them also have
difficulty learning words and speaking, and their language is
typically with spatial terms (Gabel, Cohen, Kotel, & Pearson, 2013).
Children with severe autism (AU) are not interested in
communicating; consequently, they lack the abilities needed to
effectively initiate, maintain, and end a reciprocal interaction. This
limits their opportunities to mentally build the word for social
behaviour (Low & Lee, 2011). Their language learning and
intervention outcomes therefore tend to be different from those of
children with other disabilities. On the children with a severe
physical disability (PD), their mobility is restricted and they have
restricted in participating in activities (Florian et al., 2006).
Moreover, students with different disability severities in different
education systems do not use the same abilities in their school
adaptation. It is obvious that the abilities required in an inclusive
setting may be different from those required in segregated
environments because the two education systems have different
conditions and resource types. Therefore, the present study will
examine the relationships among intervention timing, language
skills, and school adaptation individually for each disability type
and education setting.
For the students with SD studying in inclusive school, it needs
ensuring them access positive social relationship and learning
opportunities (Carter et al., 2015). The purposes of this study are to
attempt, based on the research findings, to improve current early
intervention policies and allow the most students with SD to study
and adapt well in an inclusive environment. It also seeks to facilitate
the adaptation factors if the student with SD is placed in a segregated
environment. Here are some questions this research intends to
answer: (1) Do the students with SD adapt well in inclusive
education settings or segregated settings? What factors made the
children with SD be placed in an inclusive or a segregated education

31
system? (2) How the relationship among intervention timing,
language skills, and school adaptation differs among the students
with different disability type? (3) How the relationship among
intervention timing, language skills, and school adaptation differs
between the students in inclusive and segregated educational
systems.
Method
Research Design. There are three latent variables used for
analysis in this study: intervention timing, language skills, and
school adaptation. These variables were derived from survey data
collected from the parents of Taiwanese junior high school students
with SD. These data were retrieved from the database of the Special
Needs Education Longitudinal Study of Taiwan (SNELS). In
accordance with previous studies, a number of observed variables in
the survey data which were reviewed and revised by 12 special
specialist were considered to define the three latent variables. Next,
the three latent variables were quantified by Confirmatory Factor
Analysis (CFA). The CFA model contains the three latent variables,
and each latent variable is factored by observed variables. The
following explains what each latent variable measures and the
observed variables identified via CFA in them (see Table 1).
1. Intervention timing. It refers to the time a child starts to receive
treatment to improve his/her development. This intervention must be
afforded by professionals, who are be either special educators,
therapists, or medical professionals. The observed variables of
quantifying intervention timing were the earliest age of the child
involved the intervention, the earliest age of the child’s disability was
identified, the earliest age of the child receiving a disability diagnose,
and the earliest age of the child involved special education. The first
two variables were chosen using the CFA to quantify the intervention
timing latent variable. The unit of the variables was age.
2. Language skills. They refer to the oral communicating skills in
expression and reception exhibited. The observed variables in
quantifying this latent variable included parents’ evaluations of their
kid’s language expression ability compared with peers, their kid’s
language comprehension ability compared with peers, their kid’s
verbal expression being understood by strangers compared with peers,
and their kid’s willingness to initiate language with others compared to
peers. The first three variables were determined by CFA to quantify the
latent variable. The score of the three observed variables distributed
from 1 to 4, where 1 indicated that the student’s language skills were as
good as his/her schoolmates’, 2 indicated inadequate language skills, 3
indicated poor language skills, and 4 indicated that the student cannot
communicate with others at all.
3. School adaptation. In this study, school adaptation was represented by
the children’s social and academic performance in school. The observed
variables for quantifying school adaptation included parental

32
satisfaction with their children’s interactions with teachers, interactions
with classmates, participation in activities, academic performance, and
the parents’ overall satisfaction with their kid’s school experience. The
CFA indicated that all five variables quantified the latent variable. The
score of the observed variables distributed from 1 to 4, where 1
indicated very satisfied, 2 satisfied, 3 unsatisfied, and 4 very
unsatisfied.
In CFA, the fitting observed variables are preserved in the model, and the
loading factor of each observed variable was determined to quantify the latent
variables (see Table 1). After the three latent variables were obtained, ANOVA
and correlation analyses were conducted to identify which factors influence
the choice of an inclusive or segregated school environment and how
intervention timing, language skills, and school adaptation related with one
another.
Table 1
Factor score weights from a CFA of intervention timing, language
skills, and school adaptation
Latent Variables Observed Variables Factor Score Weights
Intervention Timing Identification age 0.441
Intervention age 0.303
Language Skills Verbal expression 0.351
Language comprehension 0.475
Understood by strangers 0.307
School Adaptation Interaction with teachers 0.156
Interaction with peers 0.151
Activity participation 0.084
Academic performance 0.088
Overall education 0.197
Subjects. The subjects in SNELS were chosen with random from the
Taiwanese children with disabilities and age of 19 years or younger. The
survey data included the participants’ family background, demographic
information, medical histories, education, after-school activities, and responses
to several survey questions. The SNELS database was established in 2007 and
developed 20 survey waves from 2007 through 2012.
The data used in this study were collected in 2009 survey conducted
among the parents of 3180 junior high school students with disabilities.
Because the present study focused on students with SD, 866 subjects with SD
were included in the study. Among the 866 subjects, 519 subjects were male
and 347 were female. The subjects’ disability type profile is shown in Table 2.
Research Instrument. The SNELS data used in this research were obtained
from surveys conducted from 2008 to 2009. The SNELS team manage the
survey process, which includes questionnaire development, subjects sampling,
survey administration, survey data verification, and report the primal data in
their data bank. SNELS group is a survey organization supported by the

33
Ministry of Science and Technology of Taiwan. It comprises of 27 experts of
special educators, sociologists, survey experts, statisticians, and data analysts
etc.
Results and Discussions
Intervention Timing, Language Skills, and School Adaptation of
Students in Inclusive or Segregated Settings. Table 2 shows that 39.8% (345)
of the students with SD studied in inclusive schools or classrooms, and
60.2% (521) studied in segregated schools or classrooms. Post-hoc tests
revealed that children with severe sensory and physical disabilities were
more likely to study in inclusive environments than in segregated ones.
However, children with cognitive disabilities, including AU and mental
retardation (MR), tended to study in a segregated environment. What
factors made the children with SD be placed in an inclusive or a
segregated environment? Table 2 indicates that, with the exception of
children with VI (F=0.00, p>.05), the language skills of the children
studying in an inclusive environment were better than those of the
children in a segregated environment. The ANOVA data displayed in
Table 3 indicate that, with the exception of students with VI, the students
from each disability type in an inclusive environment had significantly
better language skills than those in a segregated one. However, the
differences in intervention timing and school adaptation between the
students in inclusive and segregated environments were insignificant
with the exception of students with VI (F=9.60, p<.001).
Considering that most of the students in segregated environments
had significantly worse language skills than those in inclusive
environments, it is interesting to note that the skills of the students with
VI in segregated environments were not significantly worse than those in
inclusive environments (see Table 2). This phenomenon can likely be
explained by their school adaptation. Table 3 shows that the students
with VI in segregated setting adapted themselves to school significantly
better than those who were in an inclusive environment. Students with
VI in an inclusive environment cannot receive visual feedback when
communicating with others and they cannot receive as much visual input
during instruction in inclusive classrooms as their classmates do. In
contrast, the students with VI in segregated environments have easily
access to alternative visual equipment or teaching materials, such as
voice basketball and Braille books. These supports helped VI students
adapt themselves better and learn more in segregated setting than the
students with VI in inclusive school environments. Therefore, the
students with VI did not benefit from their good language skill in
adaptation to school.
Table 2 also shows that 66.9% (n=111) of the students with a hearing
impairment (HI) studied in an inclusive environment, while only 33.1%
(n=32) studied in a segregated environment. The language skills of HI
students in an inclusive environment (1.97) were worse than most of the
students with other disability types in an inclusive environment.
However, the language skills of the HI students in an inclusive setting

Vol 15 No 5 - April 2016

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Similar to Vol 15 No 5 - April 2016

Similar to Vol 15 No 5 - April 2016 (20)

More from ijlterorg

More from ijlterorg (20)

Recently uploaded

Recently uploaded (20)

Vol 15 No 5 - April 2016