EXAMINING OFF-CAMPUS STUDENTS’ SENSE OF BELONGING AND BEHAVIORS IN
A TOWN-GOWN CONTEXT
By
Erin Hundley Carter

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Higher, Adult, and Lifelong Education-Doctor of Philosophy
2017

ABSTRACT
EXAMINING OFF-CAMPUS STUDENTS’ SENSE OF BELONGING AND BEHAVIORS IN
A TOWN-GOWN CONTEXT
By
Erin Hundley Carter
The current study emerged from the need to address student behavior issues in towngown communities and the practical need to know more about off-campus students as central
actors in these behaviors. Off-campus students have long been labeled as commuter students
because of limited recognition of the known diversity within the commuter population (Dugan,
Garland, Jacoby, & Gasiorski, 2008) and an absence of town-gown scholarship focused on
college student residents of the local community (Kemp, 2013). The current study sought to
remedy the little attention given to off-campus students through the exploration of their sense of
belonging to the local community, and in turn, how sense of belonging influenced behavior in
the town-gown context.
An American college town was the town-gown context chosen for the current study
because of its capacity to showcase the unique influence the university has on the character of
the town (Gumprecht, 2008). Using a stratified random sampling, the sample (n = 645) was
drawn from the population of undergraduate students living off campus in the City of East
Lansing, Michigan. Strata were determined based on residential density levels due to an interest
in assessing how residential environments of varying densities influenced sense of belonging in
off-campus students.
The current study utilized sense of belonging and town-gown literature to guide the
selection of exogenous and intervening predictor variables relevant in the town-gown context.
These predictor variables included demographic characteristics (e.g., age, sex, race, year in

school) and density categorization along with intervening variables such as environmental
perceptions, social interactions, and behavior participation. Criterion variables included sense of
belonging and positive and negative behavior participation. Hierarchical regression analyses
were used to determine the extent these demographic and community factors influenced sense of
belonging, as well as how sense of belonging influenced participation in positive and negative
behaviors in the town-gown context.
The study concluded off-campus students were a unique cohort in the student population,
warranting future attention from scholars and practitioners. Sense of belonging was not
predicted by demographic and density characteristics, but was positively influenced by
perceptions of community (e.g., reliability, friendliness) and positive behavior participation (e.g.,
attending community events, utilizing city services). Number of roommates, semesters lived offcampus, and residential density associated with houses were positive predictors of behavior
participation, both positive and negative. Environmental and social relations constructs were
positive predictors of both positive and negative behavior participation. Sense of belonging was
a positive predictor of positive behavior participation, but was not a predictor of negative
behaviors. The discussion offers insights and direction for town-gown administrators tasked
with creating policy and practical interventions aimed at supporting off-campus students, while
also addressing the unique challenges stemming from student behaviors in the local community.

Copyright by
ERIN HUNDLEY CARTER
2017

ACKNOWLEDGEMENTS

The doctoral experience has been a journey woven together by the many relationships,
experiences, and personal accomplishments along the way. There have been stops and starts,
new skills and abilities learned, and cognitive challenges and growth throughout. Each strand of
experience and each word put to page wove together the final product of this doctoral journey,
the dissertation. Along the way, I have had teachers, encouragers, guides, and champions
contributing to my development in this academic endeavor and the cultivation of my final
research project. I am deeply thankful for the presence of all in this journey.
I wish to extend my sincerest appreciation to my dissertation chair, Dr. Matthew
Wawrzynski for his unparalleled support and guidance throughout my doctoral journey. His
patience, wise counsel, and expertise were essential to me in the completion of this doctoral
degree. I would also like to express my gratitude to members of my dissertation committee for
their advice and suggestions, Drs. Marilyn Amey, John Dirkx, and Cary Roseth. Dr. Amey
provided perspective and guidance throughout this educational endeavor and could be relied on
to provide the needed encouragement to reach the finish line. Dr. Dirkx shared valuable insight
and counsel in early stages of my research project and always offered an encouraging word. Dr.
Roseth provided thought-provoking advice and suggestions that further refined and improved my
research project. Together, my dissertation committee provided the network of support and
guidance I needed to complete my dissertation and earn my doctoral degree.
I wish to express my gratitude to my HALE cohort mates for their friendship and
colleagueship along this doctoral journey. From day one, you have helped me to be a better
student and to appreciate the learning and growth we experienced together. Thank you to my

v

many writing buddies along the way, especially Drs. Sheila Coressel and Qiana Green. To my
amazing A-Team colleagues, Drs. Tonisha Lane, Michelle Vital, and Dali Martinez, I am forever
grateful for the tremendous support, guidance, and confidence you have given me along this
journey. I look forward to our continued evolution as a circle of colleagues and friends in the
years ahead.
I am forever thankful to my town-gown family (Ginny, Pat, Suchitra, Annette, Amy,
Janet, Elinor, Olivia and others) for introducing me to the world of university-community
relations and teaching me the value of relationships and collaboration. This professional home
provided the foundation for my dissertation research. Their support was immeasurable in
keeping me grounded and connected to the practical aspects of my dissertation research.
Lastly and most importantly, I wish to give a debt of gratitude to my family. First to my
amazing parents who instilled the love of learning in my life and always encouraged me to
follow my dreams whole-heartedly. I am appreciative of the unwavering support and belief in
me. To my extended Carter family, thank you for your encouragement and support across the
miles and over the years. To my beloved family, I am forever grateful for each of you. Kyle,
thank you for being an amazing partner in this academic and professional journey, while also
being a stellar husband and father. To my kiddos, Emma and Aiden, thank you for always being
excited about my progress on my “big paper” and for being proud of your mama! I look forward
to being your biggest champion in all of life’s journeys ahead.

vi

TABLE OF CONTENTS
LIST OF TABLES ...........................................................................................................................x
LIST OF FIGURES ...................................................................................................................... xii
CHAPTER ONE ..............................................................................................................................1
Introduction to the Study .................................................................................................................1
Conceptual Framework ....................................................................................................................5
The Need to Belong: A Fundamental Human Motivation .........................................................5
Sense of Community: A Theory and a Model ...........................................................................7
A Blended Framework ...............................................................................................................8
Research Questions ........................................................................................................................10
Significance of the Research Study ...............................................................................................10
Definition of Terms........................................................................................................................12
Conclusion .....................................................................................................................................13
Document Overview ......................................................................................................................14
CHAPTER TWO ...........................................................................................................................15
Literature Review...........................................................................................................................15
Town-Gown Literature ..................................................................................................................15
The Town-Gown Relationship.................................................................................................16
Benefits and Challenges of the Town-Gown Relationship ......................................................17
Benefits of the town-gown relationship .............................................................................17
Challenges of the town-gown relationship ........................................................................18
Student Behavior Concerns......................................................................................................20
Context of Town-Gown Scenarios...........................................................................................21
University Response to Town-Gown Complaints ...................................................................24
Summary of Town-Gown Literature .......................................................................................27
Commuter Literature ......................................................................................................................27
Off-Campus Students ...............................................................................................................28
Definition and Diversity of the Commuter Population ............................................................29
Challenges and Unique Needs of Commuters .........................................................................30
Need for Further Understanding of the Commuter Population ...............................................32
Summary of Commuter Literature ...........................................................................................33
Conceptual Framework Literature .................................................................................................34
A Blended Framework: Need to Belong and Sense of Community ........................................34
Theory of the need to belong .............................................................................................34
Theory of sense of community ..........................................................................................35
Application of the Conceptual Framework in the Current Study ............................................37
Examining sense of belonging ...........................................................................................37
Determining factors of influence on belonging .................................................................38
Examining sense of belonging and student behaviors .......................................................39
Sense of Belonging Literature in Higher Education ................................................................39
Studies of underrepresented students .................................................................................39

vii

Studies of additional student demographics ......................................................................40
Studies of the environment ................................................................................................41
Summary of Conceptual Framework Literature ......................................................................43
Contribution of the Current Study .................................................................................................43
Summary of the Literature Review ................................................................................................44
CHAPTER THREE .......................................................................................................................46
Methodology ..................................................................................................................................46
Purpose of the Study ......................................................................................................................46
Research Questions ........................................................................................................................46
Study Design ..................................................................................................................................47
Institutional and Community Context............................................................................................48
Institutional Context.................................................................................................................48
Local Community Context.......................................................................................................49
Sample Selection............................................................................................................................51
Data Collection ..............................................................................................................................54
Instrument ......................................................................................................................................54
Data Analysis .................................................................................................................................57
Limitations of the Study.................................................................................................................57
Conclusion .....................................................................................................................................59
CHAPTER FOUR..........................................................................................................................60
Research Findings ..........................................................................................................................60
Preliminary Data Screening ...........................................................................................................60
Sample Demographics ...................................................................................................................61
Construct Development .................................................................................................................65
Construct Reliability and Validity .................................................................................................66
Background Characteristics .....................................................................................................66
Sense of Belonging Construct..................................................................................................67
Environmental Factors Construct ............................................................................................67
Neighborhood Perceptions Construct ......................................................................................68
Neighborhood Relations Construct ..........................................................................................68
Positive and Negative Behaviors Constructs ...........................................................................69
Positive behavior construct ................................................................................................70
Negative behavior construct ..............................................................................................70
Regression Analysis .......................................................................................................................71
Evaluation of Regression Assumptions .........................................................................................72
Sample Variance Tests for Sense of Belonging .............................................................................74
Research Question One ..................................................................................................................75
Research Question Two .................................................................................................................76
Research Question Three ...............................................................................................................79
Hierarchical Regression Analysis of Positive Behaviors.........................................................80
Hierarchical Regression Analysis of Negative Behaviors .......................................................82
Summary of Significant Predictors of Student Behaviors .......................................................85
Summary ........................................................................................................................................86

viii

CHAPTER FIVE ...........................................................................................................................87
Discussion ......................................................................................................................................87
Motivation for the Current Study...................................................................................................87
Sample Profile................................................................................................................................88
Sense of Belonging in Off-Campus Students ................................................................................90
Factors Influencing Sense of Belonging ........................................................................................91
Social Factors ...........................................................................................................................91
Environmental Factor...............................................................................................................92
Demographic and Density Characteristics ...............................................................................93
Positive and Negative Behavior Factors ..................................................................................93
Exploration of Student Behaviors in a Town-Gown Context ........................................................94
Influences on Positive Behavior ..............................................................................................95
Influences on Negative Behavior .............................................................................................98
Comparing Positive and Negative Behaviors ........................................................................101
Contributions and Implications of the Current Study ..................................................................103
Contributions to Theory and Research ..................................................................................103
Contributions to Practice and Policy......................................................................................106
Recommendations for Future Research .......................................................................................111
Summary ......................................................................................................................................114
APPENDICES .............................................................................................................................116
APPENDIX A Map 1.............................................................................................................117
APPENDIX B Map 2 .............................................................................................................118
APPENDIX C Map 3 .............................................................................................................119
APPENDIX D Map 4.............................................................................................................120
APPENDIX E Participant Consent Form for Research Study...............................................121
APPENDIX F Sense of Belonging Survey Instrument..........................................................123
APPENDIX G Construct Development and Alpha Reliability for Items Used in Study ......128
APPENDIX H Correlation Matrix of Predictor and Criterion Variables ..............................131
APPENDIX I Collinearity Statistics for Final Regression Models .......................................132
APPENDIX J Sample Variance Tests for Sense of Belonging .............................................136
APPENDIX K Hierarchical Regression Analysis for Research Question One .....................138
APPENDIX L Hierarchical Regression Analysis for Research Question Two .....................139
APPENDIX M Hierarchical Regression Analysis for Research Question Three: Positive
Behaviors .........................................................................................................................140
APPENDIX N Hierarchical Regression Analysis for Research Question Three: Negative
Behaviors .........................................................................................................................141
REFERENCES ............................................................................................................................142

ix

LIST OF TABLES
Table 1 Socioeconomic Characteristics Present in American College Towns ..............................23
Table 2 Adjusted Population and Sample Frame Count at Each Stratum Level ...........................53
Table 3 Participant Demographic Characteristics for Sample and Reduced Sample ....................63
Table 4 Cronbach’s Alphas for Constructs Used in Multiple Regression Analyses .....................66
Table 5 Descriptive Statistics for Constructs (n = 589) .................................................................71
Table 6 Descriptive Statistics for Interval and Ratio Variables After List-Wise Deletion (n =
609) ..........................................................................................................................................73
Table 7 Descriptive Statistics for Interval and Ratio Variables in Final Reduced Sample (n =
589) ..........................................................................................................................................73
Table 8 Summary of Statistically Significant Predictors of Sense of Belonging ..........................79
Table 9 Summary of Statistically Significant Predictors of Positive and Negative Behaviors .....85
Table G1 Construct Development and Alpha Reliability for Items Used in Study .....................128
Table H1 Correlation Matrix of Predictor and Criterion Variables .............................................131
Table I1 Research Question One: Collinearity Statistics for Final Hierarchical Regression Model
of Background Characteristics and Density Levels on Sense of Belonginga (n=589)...........132
Table I2 Research Question Two: Collinearity Statistics for Final Hierarchical Regression Model
of Demographic, Density, Factors, Perceptions, Relations, and Behavioral variable on Sense
of Belonginga (n=589) ............................................................................................................133
Table I3 Research Question Three: Collinearity Statistics for Final Hierarchical Regression
Model of Background, Density, and Sense of Belonging on Positive Behavior Participationa
(n=589)...................................................................................................................................134
Table I4 Research Question Three: Collinearity Statistics for Final Hierarchical Regression
Model of Background, Density, and Sense of Belonging on Negative Behavior Participationa
(n=589)...................................................................................................................................135
Table J1 Sample T-test of Sense of Belonging as a Function of Sex (N = 589) .........................136
Table J2 Means, Standard Deviations, One-way Analysis of Variance on Sense of Belonging as a
Function of Race (N = 589) ...................................................................................................136

x

Table J3 Means, Standard Deviations, One-way Analysis of Variance on Sense of Belonging as a
Function of Year in School (N = 589) ...................................................................................136
Table J4 Means, Standard Deviations, One-way Analysis of Variance on Sense of Belonging as a
Function of Semesters Lived On-Campus (N = 589) ............................................................137
Table J5 Means, Standard Deviations, One-way Analysis of Variance on Sense of Belonging as a
Function of Semesters Lived Off-Campus (N = 589)............................................................137
Table J6 Means, Standard Deviations, One-way Analysis of Variance on Sense of Belonging as a
Function of Density (N = 589) ...............................................................................................137
Table K1 Research Question One: Hierarchical Regression of Demographic and Density on
Sense of Belonging (n=589) ..................................................................................................138
Table L1 Research Question Two: Hierarchical Regression of Demographics, Density, Factors,
Perceptions, Relations, and Behaviors on Sense of Belonging (n=589) ...............................139
Table M1 Research Question Threea: Hierarchical Regression of Demographics, Density,
Factors, Perceptions, Relations, and Sense of Belonging on Positive Behavior Participation
(n=589)...................................................................................................................................140
Table N1 Research Question Threeb: Hierarchical Regression of Demographics, Density,
Factors, Perceptions, Relations, and Sense of Belonging on Negative Behavior Participation
(n=589)...................................................................................................................................141

xi

LIST OF FIGURES
Figure 1. Blended Conceptual Framework ......................................................................................9
Figure 2. Venn Diagram of Predictors Associated with Positive and Negative Behaviors .........101
Figure 3. Map 1: Off-Campus Students in City of East Lansing ZIP Code ................................117
Figure 4. Map 2: Off-Campus Students in East Lansing ZIP Code with Grid Overlay ..............118
Figure 5. Map 3: Density Categories Based on Off-Campus Students in East Lansing ZIP
Code .......................................................................................................................................119
Figure 6. Map 4: Density Categories and Off-Campus Student Data in East Lansing ZIP
Code .......................................................................................................................................120

xii

CHAPTER ONE
Introduction to the Study
Undergraduate students, who reside off-campus in the local community where they attend
college, are at the center of tensions between local municipalities and higher education
institutions because of their association with causing numerous problems in the community.
News stories, police reports, and resident complaints cite parties, noise, litter, disorderly conduct,
public urination, drug and alcohol use, vandalism, and assaults among the problematic behaviors
caused by off-campus students in the local community (McLaughlin, 2011; Twohey, 2007). An
NBCNews.com story depicted neighbors’ distress at the “loud parties, rundown student boarding
houses and trash generated by weekend melees” in a Seattle neighborhood adjacent to the
University of Washington (Blankinship, 2008, para. 2). A story in the Daily Athenaeun, the
newspaper of West Virginia University, featured incidents of “underage drinking, drag racing,
beer pong, drug dealing” and excessive noise, litter, and parties as prevalent student behavior
issues caused by off-campus students in the City of Morgantown (Cossick, 2009, para 4). Duke
University’s Community Relations Committee emphasized the presence of “persistent problems
of student parties which frequently last until 5:00 a.m. and the attendant problems of noise, trash,
abusive language, greater levels of debauchery than in past years, and excessive drinking
including underage drinking” as problematic in the Durham community (Moyen, 2006, para. 4).
As the news stories above depict, the presence of off-campus students living in the local
community can be a significant source of town-gown tensions between a university and
community. The term “town-gown” refers to the relationship between the university, or “gown”,
and the lay people of the community, or “town”, and is historically utilized to imply a tension or

1

conflict between the entities (Mayfield, 2001; Thelin, 2011). In the history of higher education,
student behavior issues have been a prevalent part of the town-gown relationship and have been
noted for their deleterious effect on local communities and universities. The pervasive nature of
student behavior issues drives complaints about off-campus students and stimulates town-gown
tensions. Municipal leaders and community members implore the university to respond to offcampus student misbehaviors with judicial sanctions and educational efforts. Municipalities
demand fiscal support from the university to meet the growing costs of police, fire, and
emergency services dispatched in response to off-campus student behaviors (Baker-Minkel,
Moody, & Keiser, 2004). Universities face negative media coverage of student behavior issues,
which reflects poorly on the university’s reputation and fuels negative public opinion of its
college students (McLaughlin, 2011; Nichols 1990). Universities are pressured to answer
questions about the university responsibility for and response to students’ off-campus behaviors
(Hintz, 2011; McLaughlin, 2011). The interdependent nature of the town-gown relationship
compels universities and municipalities to respond and seek resolve to tensions resulting from
student behaviors in the local community.
Given these tensions, many universities invest time and resources into town-gown
relations in attempt to address problematic off-campus student issues, to lessen negative
publicity and public opinion, and to demonstrate collaborative partnership with the local
community. However, a significant problem higher education administrators confront is little is
known about the student population that resides off-campus locally and is at the center of towngown tensions. Within higher education, researchers have not previously studied off-campus
students as a distinct group within the student population but instead have encompassed them in
the commuter student population. This inclusion of off-campus students in the commuter student

2

definition is based on the shared commonality that they do not live in institution-owned housing
(Jacoby, 1989; Stewart & Rue, 1983). The broad simplicity of the commuter definition portrays
commuters as a homogeneous group based on the single criterion of residence. This portrayal
negates the diversity inherent within the population and leads to mistaken conceptualizations of
commuters as either “townies” in reference to place of residence with their parents or as
nontraditional students disinterested in the college experience because of competing life roles
such as employment and family commitments (Jacoby, 2000a; Stewart & Rue, 1983).
The inclusion of off-campus students in the commuter population is problematic because
off-campus students are not accurately reflected in the homogeneous concept of commuters as
townies or nontraditional students. Instead, off-campus students are traditional college age
undergraduate students, between the ages of 18 to 22 years old, who have lived in residence halls
before moving off-campus to the local community and are known to rent, live independent of
guardians or direct supervision, and live in group settings among dense enclaves of fellow
students (Gumprecht, 2003; Hintz, 2011; McLaughlin, 2011). These descriptive characteristics
demonstrate the underlying concern that off-campus students are not accurately reflected within
the commuter population. Furthermore, continued use of the homogeneous commuter definition
disregards the known diversity within the commuter population and fails to address the lack of
information and understanding of the off-campus student population.
The absence of insight into the off-campus student population is concerning for several
reasons. First, failing to recognize the diversity inherent within the commuter population, such
as the case with off-campus students, perpetuates an assumption college experiences and
interventions have the same effect on all students, regardless of individual differences, rather
than recognizing the potential the magnitude and effect of experiences may differ for students

3

based on differing characteristics (Pascarella, 2006). As a result, higher education practitioners
approach town-gown efforts guided by what is known about the commuter population broadly,
instead of with insight about how the off-campus student population may be similar or distinct
from commuters. Second, the Council for the Advancement of Standards in Higher Education
(CAS) (2012) implores higher education institutions to provide all students “equitable access to
institutional services, engagement opportunities, and the total educational process regardless of
place of residence” (p. 1). Striving to offer all students equitable access to educational
opportunities requires institutions to move beyond a continued reliance on the homogeneous
commuter concept and instead recognize within-group differences may exist within the
commuter population (Dugan, Garland, Jacoby, & Gasiorski, 2008). Lastly, institutional efforts
to address town-gown tensions caused by student behavior issues are weakened by the lack of
knowledge of the off-campus student population, including limited understanding of the
demographic composition of this group, minimal insight into the nature of the collegiate
experience from a local place of residence, and little knowledge of student involvement and
behaviors beyond what is known about town-gown tensions.
Cumulatively, the lack of information about off-campus students limits efforts to support
these students adequately and hampers efforts to address behavior issues central to town-gown
concerns. Thus, the current study was focused on examining off-campus students as a group
within the student population, with emphasis on gaining insight into the nature of their
experience and behavioral choices within the off-campus community. The study aimed to
broaden understanding of the demographic composition of the off-campus student population,
offer insight into the behavioral choices of off-campus students in the local community, and
examine potential factors influencing feelings of belonging to the local community.

4

Conceptual Framework
The purpose of a conceptual framework is to provide a way to think about and organize
ideas to achieve a research purpose. In the current study, the purpose was to examine offcampus students and their behavioral choices in the local community as a means of gaining
insight on student behavior concerns central to town-gown tensions. Given this purpose, the
conceptual framing of this study blended theoretical components from two distinct, but related
concepts, namely the theory of the need to belong (Baumeister & Leary, 1995) and the theory of
sense of community (McMillan & Chavis, 1986). The following section details each theory and
explores their use as the conceptual framework for the current study.
The Need to Belong: A Fundamental Human Motivation
The theory of the need to belong (Baumeister & Leary, 1995) served as the primary
conceptual underpinning for the current study because the need to belong provided a theoretical
construct to examine students’ motivations in the town-gown environment. The usefulness of
the need to belong as a construct to study motivation stemmed from its designation as a
fundamental human motivation. On a theoretical level, Baumeister and Leary (1995) established
the significance of the need to belong by arguing for the need to belong to be designated as a
fundamental human motivation. The authors successfully demonstrated the need to belong
satisfied criteria inherent in a fundamental human motivation, including producing effects in a
spectrum of settings, yielding affective and cognitive responses, eliciting goal-oriented behavior,
leading to ill effects (e.g. on health or adjustment) when thwarted, being universally applicable to
all people, and affecting a wide range of behaviors. Framed as a human motivation, the authors
argued the need to belong is innate, universal among humans, and a natural driver toward
seeking and sustaining belonging. As such, humans have an inherent need for belonging and the

5

pervasive drive to satisfy this need is sufficient to motivate behavior.
The robustness of the need to belong construct (Baumeister & Leary, 1995) has led to its
conceptualization and use within higher education contexts and populations as the construct of
sense of belonging. Hurtado and Carter (1997) among others, used the sense of belonging
construct as an alternative conceptualization for the concept of integration within higher
education persistence models (Tinto, 1993). Use of the sense of belonging construct addressed
the concern models of persistence focus on behavioral aspects of integration while neglecting the
psychological need to belong is inherently human (Hausmann, Ye, Shofield, & Woods, 2009).
Similarly, Strayhorn (2012) asserted sense of belonging refers to a student’s perceived social
support and feelings of mattering and connection to others in college. Satisfying this need to
belong leads to positive, prosocial, and productive outcomes in education including achievement,
engagement, optimal functioning, persistence, retention, and well-being (Hausmann et al., 2009;
Rhee, 2008; Strayhorn, 2012).
Given the positive outcomes associated with achieving a sense of belonging in college,
the belonging construct has gained greater appreciation and use within higher education research.
Researchers have used sense of belonging to understand how belonging influences the college
experience of students of Color (Hausmann et al., 2009; Hurtado & Carter, 1997; Johnson et al.,
2007; Strayhorn, 2008; Strayhorn, 2012), first-year students (Hoffman, Richmond, Morrow, &
Solomone, 2002), graduate students (Strayhorn, 2012), and part-time students (Kember, Lee, &
Li, 2001). Additionally, the role of the environment in shaping sense of belonging was explored
by Freeman, Anderman, and Jensen (2007) examining classroom and campus belonging and
Johnson et al. (2007) examining residence hall environments.
In the current study, the sense of belonging construct provided a guide to consider how

6

the desire to belong may serve to motivate off-campus students’ behavior choices in the local
community. To further explain, suppose off-campus students have a desire to belong to the
community they reside in and by striving for this belonging they participated in normative
behaviors of the community, as do other community members. Within the context of the towngown environment, these normative behaviors include the student behaviors central to towngown tensions (e.g., partying, alcohol consumption) and typical neighborhood behaviors (e.g.,
yard maintenance, neighbor interactions). Engaging in these normative behaviors becomes a
means for off-campus students to gain a sense of belonging within the community. Thus, in the
current study, the need to belong served as a useful concept to examine how off-campus
students’ desire for belonging may be sufficient motivation to act in pursuit of achieving and
maintaining a sense of belonging to the community.
With the emphasis on belonging in a town-gown context, the current study expanded
belonging literature to include a context external to the university, namely the local community
students reside in off campus while attending college. Given sense of belonging had not
previously been utilized to study contexts external to the university, an additional construct to
guide the examination of belonging to the local community was needed. Thus, the theory of
sense of community was selected to compliment the sense of belonging construct and enhance
the conceptual framework for the current study.
Sense of Community: A Theory and a Model
The concept of sense of community was closely related to the notion of sense of
belonging, as both pertained to the interpersonal connections and affiliations individuals desire
and feel in relation to others. McMillan and Chavis (1986) defined sense of community as “a
feeling that members have of belonging, a feeling that members matter to one another and to the

7

group, and a shared faith that members’ needs will be met through their commitment to be
together” (p. 9). McMillan and Chavis (1986) identified four elements essential in the
cultivation of sense of community including a perception of membership, influence, integration
and fulfillment of needs, and shared emotional connection. The authors emphasized these
elements reflect the intimate, dynamic nature of development and maintenance of a sense of
community.
Based on the theory of sense of community (McMillan & Chavis, 1986), Chavis and
Wandersman (1990) created an empirical model to reflect the generation of sense of community.
In the model, sense of community acted as a catalyst by mobilizing key variables to motivate
participation in local action, specifically participation in voluntary neighborhood associations.
The key variables found to influence participation and belonging related to an individual’s
perception of the environment, social relations, and perceived control and empowerment within
the community. The relationship between variables in the model was found to be bidirectional
and transactional, which reflected and reinforced the dynamic nature of the community
development process. The model of sense of community (MSC) (Chavis & Wandersman, 1990)
offered a framework for examining which factors can influence the sense of belonging and
affiliation one feels toward a community. The MSC factors used in the current study included
questions related to the perception of the environment and social relations themes. The third set
of questions, namely the perceived control and empowerment theme, was not used because the
focus of the study did not extend to understanding student agency or empowerment.
A Blended Framework
In the design of the current study, the theory of the need to belong and the model of sense
of community provided the conceptual components needed to examine what influenced and

8

motivated student behaviors in the local community. The theory of the need to belong
(Baumeister & Leary, 1995) provided the belonging construct as a way to conceptualize what
served to motivate behavior in off-campus students. The model of sense of community (MSC)
(Chavis & Wandersman, 1990) provided variables to consider for their ability to influence sense
of belonging to a community or group. Figure 1 depicts the conceptual framework resulting
from the use of components from both theories.

Figure 1. Blended Conceptual Framework
In the construction of this conceptual framework, there was merit in thinking the
belonging construct, as captured in the need to belong, could provide insight into student
behavioral choices because the need to belong was known as a fundamental human desire
capable of motivating behavior. Use of the belonging construct provided an opportunity to
examine how belonging influenced positive and negative behaviors within the community.
Insight into the role belonging played in motivating behaviors was needed for designing
educational approaches aimed at increasing positive community investment while decreasing the
occurrence of negative behaviors central to town-gown tensions.
Given the motivating role sense of belonging has had on behaviors, there was worth in

9

also examining factors that influenced an individual’s sense of belonging. The model of sense of
community guided the selection of variables to explore for their ability to influence sense of
belonging to a community. Examining these variables created the opportunity to ascertain the
influence each variable had on belonging and whether the nature of the influence was positive or
negative, that is contributed to or detracted from, sense of community. With understanding of
the effectiveness and nature of influence associated with each variable, there is opportunity to be
intentional with future efforts and strategies to influence sense of belonging in community.
Research Questions
With a research study focused on examining off-campus students and their behavioral
choices in the local community, the theories of the need to belong (Baumeister & Leary, 1995)
and sense of community (McMillan & Chavis, 1986) guided the conceptual design of the current
research. Given this purpose, the following research questions steered the current study:
1. What is off-campus students’ sense of belonging to the local community? And, to what
extent do demographic characteristics and residential density contribute to the prediction
of off-campus students’ sense of belonging to the local community?
2. To what extent do environmental, social, and behavioral factors contribute to the
prediction of off-campus students’ sense of belonging to the local community?
3. What is the relationship between sense of belonging in off-campus students and their
involvement in positive behaviors (e.g., neighborhood involvement, responsible
neighbor) and negative behaviors (e.g., alcohol violations, litter, fights) within the local
community?
Significance of the Research Study
The current study sought to advance higher education research and scholarship in several

10

ways. The study aimed to provide insight into the off-campus student population as a population
of students not previously studied empirically in higher education. Instead, off-campus students
have historically been included in the commuter student population, which is problematic
because they are not accurately reflected in the broad, homogeneous definition of commuters.
Additionally, the focus on off-campus students adds to belonging literature in higher education
because off-campus students have not been studied previously with regard to sense of belonging.
Existing sense of belonging literature in higher education has also focused on examining the
relationship of belongingness toward or within the university and associated college community
(Strayhorn, 2012). In the current study, examining students’ sense of belonging to the local
community opened inquiry into the role feelings of belonging to external entities beyond the
university may play in the collegiate experience of students.
In the town-gown context, the review of relevant literature revealed minimal research
addressing the town-gown relationship between a university and community. Although a small
amount, most existing literature focused on the role of universities and communities in
addressing benefits and challenges of the town-gown relationship including economic
development, public service demands, resource allocation, and cultural and intellectual benefits
(Mullins & Gilderbloom, 2002; Nichols, 1990). A limited number of studies examined the
experience and feelings of individuals in and toward the town-gown environment, including
university administrators, city administrators, and non-student residents (Bruning, McGrew, &
Cooper, 2006; Kittle, 2005). McLaughlin (2011) examined the effect of a specific community
building program for off-campus students and its influence on neighborhood outcomes. The
current study added to the town-gown literature by focusing on students and bringing to light the
sense of belonging these students feel toward their local community.

11

A compliment to the research contributions of the current study was the practical
additions offered to administrators engaged in town-gown relations work on behalf of the
university and the local community. Insight into the sense of belonging of off-campus students
provided additional understanding of the nature of the off-campus living experience and
suggested ways practitioners can enhance services and resources designed to support this
population. As important, understanding how sense of belonging influenced behaviors in offcampus students provided direction for addressing student behavior issues and lessening towngown tensions.
Definition of Terms
To aid in better understanding the context of the current study, defining the key terms
was important. At the center of the current study was the student population captured in the
term, off-campus students. As a term, off-campus student describes an undergraduate student
who lives off-campus in the local community where the institution of higher education is located.
Off-campus students typically reside in neighborhoods or geographic areas characterized by a
large density of student residents and may also include non-student neighbors (McLaughlin,
2011). These students are typically of traditional college age (i.e., 18-22 years) and most have
lived on-campus in residence halls at least one year in compliance with a first-year student
residency requirement (Hintz, 2011). An off-campus student is likely to live in a rental property,
may live alone or with peers, and may walk or drive to campus (Kuh, 2003; Kuh, Gonyea, &
Palmer, 2001). Typically, off-campus students have limited experience living off-campus with
minimal oversight and are unfamiliar with the guidelines, expectations, and policies associated
with community living (Hintz, 2011). As a result, off-campus students are at the heart of towngown tensions due to student behaviors locally.

12

Off-campus students have traditionally been subsumed under the term commuter student
or commuters in existing literature. The term commuter students or commuters is used to
describe undergraduate students who do not live in institution-owned housing. As a term,
commuter encompasses a broad diversity of students and does not differentiate between any
subgroup differences in the population (Dugan et al., 2008). Traditional and non-traditional age,
full and part-time enrollment, and commuting distance to campus are just a few descriptors used
to portray the commuter population (Jacoby, 1989, 2000a).
The presence of off-campus students in the local community has caused an increased
focus on and effort dedicated to town-gown relations. The term, town-gown is used to describe
the relationship between “institutions of higher learning and the communities in which they
reside” (International Town-Gown Association, 2013a). Universities and communities have a
significant influence on each other, both directly and indirectly, in ways beneficial and
detrimental to the health of the shared university community. Positive effects include economic
and cultural benefits, educational opportunities afforded, and shared resources and opportunities
across entities (Nichols, 1990). Negative effects stem from power struggles between leadership
of both universities and communities; resource disparity toward infrastructure demands related to
parking, road maintenance, and safety; and student behaviors associated with parties, alcohol
consumption, and littering cause quality of life concerns for the community (Nichols, 1990).
The current study focused specifically on the student experience in the town-gown context.
Conclusion
Chapter 1 served to introduce the current study, which focused on town-gown tensions
between a university and community resulting from student behavior concerns attributed to offcampus students. Compelled to alleviate town-gown tensions, institutions discover off-campus

13

students have been subsumed within the broader commuter population and have not yet been
studied as an independent group. The current study sought to address this knowledge gap by
examining the off-campus student population, their sense of belonging to the local community
where they reside, and factors influencing belonging. Utilizing sense of belonging provided an
opportunity to examine how belonging motivates behavioral choices and in turn, how cultivating
belonging may help alleviate behavior issues central to town-gown tensions.
Document Overview
Chapter 1 provided an overview of the research problem and questions guiding the
current study of off-campus students and their behavioral choices within the local community, as
well as explained the significance of the study in this area. Chapter 2 explores literature
pertinent to the current study, namely in the areas of town-gown relations, the commuter student
population, and sense of belonging. Chapter 3 details the methodology of the current study.
Chapter 4 organizes the findings of the study around the research questions guiding the study.
Chapter 5 discusses the results of the current study and offers implications for research and
practice in higher education.

14

CHAPTER TWO
Literature Review
Off-campus students have been largely absent from higher education literature and as a
result, little is known about the nature of their experience from a place of residence in the local
community. This limited knowledge is concerning because off-campus students are embroiled in
town-gown tensions and information is needed to better support these students and to begin to
address the tensions at hand. This literature review was designed to situate the current study
within existing literature, thus this chapter is divided into three parts reflecting literature on the
town-gown relationship, commuter students, and the need to belong. The first section focuses on
town-gown literature by exploring student behavior concerns in the context of the town-gown
relationship and the challenges and benefits inherent within. The second section provides an
overview of the commuter population including its diversity, limitations of the commuter
definition, challenges and unique needs, and justification for further study of commuters. The
third section examines belonging literature to provide greater insight into the theoretical
construct framing this study and to shed light on existing belonging research in higher education.
Town-Gown Literature
The initial section of the literature review serves to introduce and explore the innate
relationship in town-gown interactions between a university and local community. The review
depicts the benefits and challenges inherent within the town-gown relationship and explores the
tensions elicited by student behavior issues in the community. The context of town-gown
relationships is examined as population size and characteristics of the university and local
community can influence the significance of the town-gown relationship and the salience of the
tensions caused by off-campus students. The section concludes with a focus on efforts used by

15

universities to respond to student behavior concerns, which are central to town-gown tensions.
The review of town-gown literature is essential in establishing an understanding of the
context and relationship that make off-campus students a population of significant interest to
both the local community and the university, and thus the population of focus in the current
study. Furthermore, an understanding of the interdependent nature of the town-gown
relationship provides a rationale for why universities invest fiscal and human resources in towngown work related to off-campus students, as a response to existing tensions, and a responsibility
to the long-term health of the town-gown relationship.
The Town-Gown Relationship
The term “town-gown” refers to the distinct spheres differentiating the university, or
“gown”, from the lay people of the community, or “town” (Mayfield, 2001; Thelin, 2004).
Dating back to medieval Europe, the town-gown term highlights the distinction between the
university and the community and implies a tension or conflict between them. As Kemp (2013)
explained, the “misunderstandings between these parties can be traced to historical developments
and a lack of general understanding of the implications that these developments have on each
other” (p. 1). The history of interactions between the university and local community is marked
with issues evolved from separate governing bodies, competing priorities and loyalties, and
limitations of a shared physical geography (Baker-Minkel et al., 2004).
Despite the challenges associated with the town-gown relationship, the university and
community are linked in an interdependent relationship where the well-being of one entity
affects the well-being of the other and only at their own peril do they ignore one another or act
alone (Freeland, 2005). The interdependent relationship is evident as the university contributes
to the economic vitality of the community through its academic and operating activities and in

16

turn, the economically vibrant community demonstrates the viability of the region and helps the
institution attract and retain high-quality students, faculty, and opportunities (Porter & Grogan,
2002/2013). Together, as Porter and Grogan (2002/2013) explained “greater economic vibrancy
and more successful academic institutions will in turn contribute to the competitiveness of the
broader urban and regional economy” (p. 223). The significance of this interdependence was
evident in the resilience of town-gown environments following the 2009 financial crash in the
United States (Florida, 2013). Despite this financial turbulence, “college town” environments
experienced low unemployment rates, high-paying job markets, growth in entrepreneurial startups, and a stable economy when compared to communities without the presence of a higher
education institution (Florida, 2013). The well-being of the university and the local community
are linked for the betterment of each entity individually and the shared environment.
Benefits and Challenges of the Town-Gown Relationship
The review next explores the benefits and challenges inherent in the interdependent
relationship between the university and the community. The shared commitment toward the
well-being of the town-gown relationship drives both the university and the community to work
toward promoting the health of the relationship and to address challenges that arise.
Benefits of the town-gown relationship. The interdependent nature of the town-gown
relationship yields positive effects for the university and local community. As a beacon of
academic enterprise, a university provides the local community with opportunities for
educational enrichment, cultural experiences, and entertainment (Baker-Minkel et al., 2004;
Warfield, 1995). The university attracts a diverse, youthful student population to the region
which fosters an energetic, dynamic, and cosmopolitan atmosphere attractive to residents and
visitors (Gumprecht, 2003). Students actively fulfill employment, internships, and volunteer

17

needs for local businesses, schools, and service agencies (O’Mara, 2012). The university
provides economic benefits including local spending by campus constituents, visitors, and the
university itself (Baker-Minkel et al., 2004), serving as a major employer within the geographic
region (Nichols, 1990), and fostering the retention of an educated workforce (O’Mara, 2012).
University involvement in research and technology transfer yields regional efforts of innovation
and entrepreneurship, contributes to talent retention of educated individuals, and attracts top-tier
businesses to the region (O’Mara, 2012; Sungu-Eryilmaz, 2009/2013). Together, these benefits
contribute to the economic stability and viability of the local community and foster the success
and reputation of the university.
The university provides many benefits to the local community and the local community
reciprocates. Local municipalities provide infrastructure and public services crucial to university
functioning, such as water and sewer services, road maintenance, and public safety services
including police and fire departments (Nichols, 1990). Local communities offer amenities such
as shopping, dining, arts, and entertainment that are attractive to university constituents and
showcase the viability of the region to potential students, faculty, and staff (O’Mara, 2012).
Municipal neighborhoods provide viable housing options for faculty, staff, and students. For
students, the local community can be a source of employment with local businesses, for
engagement in service-learning experiences, and for placement with student teaching
assignments or internships (O’Mara, 2012). An aesthetically pleasing, safe, and vibrant local
community serves as a recruitment tool for universities seeking to attract talented students,
faculty, and staff to the institution and the region.
Challenges of the town-gown relationship. Despite benefits accrued in a town-gown
relationship, the relationship presents challenges and is often marred by conflict and tension.

18

One primary tension stems from financial concerns. As tax-exempt institutions, universities are
not required to contribute financially to local municipalities, although some universities provide
payments in-lieu of taxes or on a project-by-project basis to offset the cost of municipal services
provided to the university (Baker-Minkel et al., 2004). Despite these ad hoc arrangements,
municipalities face dwindling financial resources from state and federal sources while
experiencing increased costs associated with land use projects and infrastructure demands.
Instead of providing financial support, universities “respond by emphasizing the institution’s
positive impact on nearby property values, local business activity, and numerous other economic
benefits not captured directly in the city budget” (Baker-Minkel et al., 2004, p. 8). This
university response fuels tensions and results in quid pro quo arrangements for zoning and
development projects instead of mutually beneficial actions (Freeland, 2005).
Beyond economic tensions, frequent complaints emerge related to infrastructure and
public service challenges. During the academic year, the university presence is felt in the traffic
and parking congestion present on the roadways due to the high volume of students, faculty, and
staff traveling throughout the campus and community. Increased demand on roads and streets
affects street maintenance departments and police services and adds financial and resource
burdens to city budgets (Nichols, 1990). Police, fire, and emergency medical services are in high
demand covering routine responses to campus calls and special events like major football games
and entertainment events (Martin & Samels, 2006/2013). Universities are frequently blamed for
the behavior of students in the local community including parties, noise, and drunken conduct
and in turn, are expected to assist with the additional cost of police, fire, and public service
staffing needed to respond (Baker-Minkel et al., 2004). With the extra demand on city services,
municipalities often seek to assess a fee or request additional funding from universities for

19

infrastructure and resource needs. Universities experience this request as a “squeezing” for funds
instead of a collaborative effort. University and community negotiations regarding human,
fiscal, and resource constraints perpetuate a problematic tone and portray a politically
challenging environment instead of nurturing the town-gown relationship as an asset.
Student Behavior Concerns
Off-campus students are a vibrant and dynamic part of the town-gown environment as
they bring a youthful energy and enthusiasm to the community and contribute as valuable,
involved citizens, neighbors, volunteers, and employees. Despite the benefits associated with
off-campus students, the persistent presence of negative off-campus behaviors is a central driver
of tensions in the town-gown environment and often overshadows the positive contributions of
off-campus students. Universities are motivated to resolve these student behavior issues because
of their commitment to the town-gown relationship and their investment in helping students to
develop as responsible, engaged citizens. Common complaints cited by city staff and neighbors
include: excessive alcohol consumption, parties, noise, garbage, litter, parking, over-occupancy
in rental property, and acts of vandalism, personal injuries, and disorder (Harasta, 2008;
Hubbard, 2009; Munro, Turok, & Livingston, 2009; Young, 2002/2013).
Student behavior issues are predominant in the town-gown relationship in part because of
the proliferation of off-campus rental housing in the local community. Over time, off-campus
housing has grown as residence hall capacities have not kept pace with growing college
enrollments (Munro et al., 2009) and students frequently desire independent living away from
university oversight (Nichols, 1990). Because students desire to live near campus, many
adjacent residential neighborhoods have been transformed as single family homes are purchased
by landlords and converted into multi-unit dwellings for student renters. The influx of rental

20

properties and student occupants, as Nichols (1990) explains, has caused “deteriorating
neighborhoods, contrasts and conflicts with the life-styles of permanent residents, declining
property values, and zoning nightmares for city governments” (p. 11). The increased student
presence causes noticeable social, cultural, physical, and economic changes to the local area, a
process Munro et al. (2009) defined as “studentification” (p.1807). As a change process
occurring over time and not as a once and for all adjustment (Munro et al., 2009),
“studentification” speaks to the importance of ongoing attention devoted to the evolving
presence of off-campus students in the community.
Context of Town-Gown Scenarios
The salience of the concerns generated by student behavior issues in a local community
can vary, in part, because of the town-gown context in which the aggressions occur. Although
the town-gown relationship is invariably present to some degree in all municipalities in which
institutions of higher education are located, the influence of the university on its community is
measured in part by its size in relation to the city or town it inhabits. As the university
population increases in relation to the city population, Nichols (1990) explained “the greater the
impact that the institution will have on the city’s economy, social life-style, cultural
environment, and physical environment” (p. 15). In circumstances where the student enrollment
comes close to matching or exceeding the city population, the town-gown relationship is
operationalized at an organizational level across the institution and city-government and has
significant influence on both entities. In contrast, institutions in large urban areas have little
influence on the overall city environment and instead the town-gown relationship is based on
neighborhood areas geographically influenced by the university presence (Nichols, 1990).
American college towns are a salient example of the significance the town-gown

21

relationship can take on when the size of the institution creates a considerable influence on the
local community. Gumprecht (2008) defined an American college town as “any city where a
college or university and the cultures it creates exert a dominant influence over the character of
the town” (p. 1). Similar to one another, college towns differ from other cities and regions in
fundamental ways because the influence of the collegiate culture is concentrated, conspicuous,
and clearly dominant. Using socioeconomic characteristics, Gumprecht (2008) confirmed
college towns were fundamentally different from other cities across the United States. The
socioeconomic characteristics Gumprecht (2008) found to be uniquely present in college towns
are summarized in Table 1.
The socioeconomic characteristics described by Gumprecht (2008) help explain the
context making college towns unique and salient environments to observe the fullness of the
town-gown relationship. Factors such as the presence of a youthful population, the transient
nature of the population from year to year, and the higher incidence of rental and roommate
living arrangements demonstrate the influence of the student population on the local community
and illuminates the incongruence of these traits with other characteristics found in American
college towns, including higher affluence and high cost of living. Munro et al. (2009)
emphasized “students have distinctive demographic and economic characteristics disproportionately young, middle class, with no dependents, and well educated - which broadly
remain stable as each cohort replaces its predecessor” (p. 1807). The concentrated influence of
college students on the community was significant and lasting, which made college town
environments ideal for the study of off-campus students within the town-gown community.

22

Table 1
Socioeconomic Characteristics Present in American College Towns
Socioeconomic characteristic

Summary description of characteristic

Youthful population

Average median age for college towns in the study was 10
years younger than median age for U.S. On average, more
than one third of the town population was 18-24 years old.

Highly educated population

Population was twice as likely to hold a college degree, six
times more likely to hold a doctorate than U.S. population.

Residents are more likely to
work in white-collar jobs

Few college towns possess heavy industry and as a result the
population was half as likely to work in a manufacturing job
than the U.S. Instead, the population worked in managerial
and professional jobs (50%) and in education (25%).

Comparatively affluent
community

Towns are prosperous, unusually economically stable, and
have low unemployment rates. Median family income
averaged $10,000 higher than similar places nationwide.

Cost of living is high

Overall cost of living averaged 2.4% higher than U.S., while
housing costs averaged 4.5% higher.

Transient place

The population and pulse of a college town rose and fell
with academic calendar. Students moved to college as first
years, moved between years, and moved away at graduation.

Residents are more likely to
rent, live in apartments, and
have roommates

Residents were less likely to own homes and more likely to
live in multi−unit dwellings or group housing. More than
50% of the study group lived in rental property compared to
30% of the U.S. population. Specifically, 20% lived in
group or multi−unit dwellings compared to 2.8% nationally.

Towns are cosmopolitan

Colleges attract students and faculty nationally and
internationally and as a result, towns were more ethnically
and culturally diverse than comparable cities.

Towns are unconventional
places

Compared to the general U.S. population, college town
residents were less religious and more likely to: walk or bike
to work, listen to public radio, vote for left-wing political
candidates, and shop at a food cooperative.

Quality of life is high

Known for lively downtowns, beautiful residential areas,
impressive cultural opportunities, ample recreational
options, good schools, and safe streets, college towns ranked
high on lists of best places to live, retire, or start a business.

23

University Response to Town-Gown Complaints
The prevalence of student behavior concerns within the local community causes tension
in the town-gown relationship and is particularly endemic in a college town environment. The
significance of the behavioral concerns and resulting tensions drive the university to respond and
address the behavioral issues caused by off-campus students in the local community. Beyond
their own intentions to act, the university receives requests and faces pressure from external
entities to response to behavior concerns. Resident complainants call on the university to
address student behavior issues through judicial and educational avenues despite the occurrence
of the behaviors beyond university boundaries. Municipal government asks the university to
contribute to additional costs associated with police and fire response to problems and
subsequent public service demands for clean-up (Baker-Minkel et al., 2004; Martin & Samels,
2006/2013; Nichols, 1990). Journalists raise public awareness of town-gown tensions by writing
stories depicting the melee caused by student behaviors, the ensuing resident complaints, and the
action, or lack of action, taken by either the university or city leadership in response.
In response to complaints and negative publicity, the university invests time and
resources in town-gown relations with an aim toward changing public opinion and demonstrating
engagement as a collaborative partner with the local community. The Town/Gown Task Force
of the University of Florida and City of Gainesville (Young, 2002/2013) and the joint
Neighborhood Relations Committee of the City of Williamsburg and the College of William and
Mary (Whitson, 2011/2013) exemplify the commitment of universities to town-gown relations
with stated goals of being good community partners, improving relationships between entities,
and improving the quality of life for all in the town-gown environment. The Community
Relations Coalition serves to “build trusting relationships between the communities of East

24

Lansing and Michigan State University by fostering cooperation and communication between
neighbors” (Community Relations Coalition, 2013, About us section). The International TownGown Association (ITGA, 2013a) encourages “civic leaders, university officials, faculty,
neighborhood residents and students to collaborate on common services, programs, academic
research and citizen issues, creating an improved quality of life for all residents, students,
visitors, faculty and staff” (Mission and Vision section, para. 2). Similarity in these stated goals
demonstrates the consistency of purpose in town-gown relations across multiple university and
community contexts.
Universities employ a variety of strategies and approaches in efforts to address towngown tensions caused by students. Although some approaches are punitive, such as severe
penalties and additional police interventions, many more efforts are conciliatory and educational,
such as communication campaigns aimed at students (Kaplowitz & Campo, 2004). Great
commonality exists in the types of educational programs offered by town-gown offices in the
United States as most feature informational campaigns about being good neighbors, partying
smart, and raising awareness of city ordinances, neighborhood responsibilities, and city services
(Colorado State University [CSU], 2014; Michigan State University [MSU], 2014; The Ohio
State University [OSU], 2014; University of Colorado Boulder [UCB], 2011; University of
Florida [UF], 2013). Many offices offer conflict resolution and mediation services for students
in need of addressing off-campus concerns with neighbors, be they student or non-student
residents (CSU, 2014; Whitson, 2011/2013).
Dissemination of educational campaign materials occurs via print publications (e.g.,
flyers, booklets), robust website platforms, email communications, door-to-door outreach
conducted within neighborhoods, and neighborhood events, such as barbecues and ice cream

25

socials (CSU, 2014; MSU, 2014; OSU, 2014; UCB, 2011; UF, 2013; Whitson, 2011/2013).
Educational efforts are often collaborative involving university and city staff, community
members, students, and property managers in the creation and dissemination of information
through various avenues. Many programs use off-campus students as peer educators in towngown outreach to serve as knowledgeable resources, communication disseminators, and
community builders among students and community members (Hintz, 2011).
In other cases, coalitions or committees, such as the Community Relations Coalition of
Michigan State University and the City of East Lansing, the Joint Neighborhood Relations
Committee of the City of Williamsburg and the College of William and Mary, and the
Town/Gown Task Force of the University of Florida and the City of Gainesville, were formed to
involve multiple stakeholders in the responsibility for town-gown efforts. The hallmark of
coalition structures is their strategic design as umbrella entities where the university, city, and
other stakeholders share joint responsibility and accountability for town-gown efforts (MSU,
1999). The mutuality of the arrangement ensures ongoing investment and engagement by all
entities involved.
A common philosophy serves as a foundational guide for the many strategies and
interventions used across the nation to address town-gown tensions caused by students. By and
large, efforts balance between increasing awareness of community regulations and expectations,
while promoting increasing interactions and rapport building among student and non-student
neighbors. Strategies such as neighborhood events and “be a good neighbor” campaigns are
intentionally designed to establish channels of communication between residents and to build a
sense of community. As McLaughlin (2011) explained, the underlying goal driving these efforts
is to “build stronger, more educated, and self-regulating communities” (p. 10) with an end

26

towards cultivating a sense of community, promoting greater harmony among neighbors, and
ultimately decreasing town-gown tensions. Efforts are aimed at cultivating a sense of belonging
to the community with an intention positive community outcomes around behavior and
participation will be elicited. Ultimately, the goal is for individuals to become more invested and
engaged in the local community, leading them to be more respectful and responsible citizens.
Summary of Town-Gown Literature
The town-gown literature review has demonstrated the dynamic, interdependent nature of
the relationship between a university and local community. Benefits of the town-gown
relationship range from economic stability to expansive cultural, social, and educational
opportunities, while challenges stem from infrastructure demands, resource challenges, and
student behavior complaints. Student behavior issues emerged as a consistent complaint among
stakeholders and served as a central focus of university attention toward and investment in
repairing town-gown tensions. The link between geographic context and significance of the
influence of student behavior issues on the town-gown relationship was examined and revealed
the salience of the American college town environment for study of student behavior issues. In
the university response to student behavior concerns, the notion of cultivating a greater sense of
community belonging to elicit positive community outcomes emerged as an underlying driver
influencing town-gown efforts. Next, the literature review turns to commuter literature, given
the current study focused on off-campus students as central actors in town-gown tensions and
off-campus students have long been categorized within the commuter population.
Commuter Literature
The review of commuter literature begins with further examination of off-campus
students as central actors within the town-gown relationship and illustrates the existing affiliation

27

with the commuter student population. Next, the commuter definition is reviewed and its
limitations are discussed considering the known diversity of the commuter population. Then, the
challenges and unique needs of commuter students are examined, followed by a call for further
understanding of the diverse commuter population, including off-campus students.
Off-Campus Students
The current study focused on undergraduate students who live off-campus in the local
community and are at the heart of town-gown tensions between universities and local
communities. Off-campus students represent an increased segment of the local community as
undergraduate enrollments have grown and university residential facilities are unable to
accommodate the increase (Munro et al., 2009). The growing presence of off-campus students in
the local community intensifies the student influence on the community dynamic and makes their
actions conspicuous and noticeable by city administrators, community residents, neighborhood
leaders, and police. Consequently, off-campus students are central to town-gown tensions
because behavior concerns such as parties, litter, noise, excessive alcohol consumption, and
disorderly conduct drive resident complaints, fuel municipal demands for resources, and elicit
negative media coverage of the university (Hintz, 2011; McLaughlin, 2011; Nichols 1990).
Universities invest resources and staffing in town-gown relations to address problematic
off-campus student behaviors and issues. However, administrators working in town-gown
relations quickly discover an absence of information about off-campus students and the nature of
their college experience from a place of residence in the local community. The limited
information known about off-campus students identifies them as being traditional college aged,
living alone or with peers in rental housing, and attending college full-time (Gumprecht, 2003,
McLaughlin, 2011). Off-campus students are known to have lived in residence halls at least one

28

year, only recently moved off-campus, and have minimal experience with living on their own
without oversight or guidance from parents or university staff (Hintz, 2011). Beyond these
descriptive characteristics, the information known about off-campus students is limited to the
disruptive behaviors central to town-gown tensions.
A primary obstacle limiting knowledge of off-campus students stems from their historical
inclusion within the commuter student population in higher education. Off-campus students
have traditionally been defined as commuter students based on the trait of not living in
institution-owned housing (Jacoby, 1989; Stewart & Rue, 1983). However, most off-campus
students are not accurately reflected in the characteristics associated with commuters because the
commuter definition is antiquated and does not reflect the known diversity within the commuter
population. As a result, further insight into the off-campus student population is needed but first
requires additional understanding of the definition and concept that has historically defined the
commuter population and included off-campus students. The next sections explore the commuter
population in greater detail beginning with the commuter definition and its limitations.
Definition and Diversity of the Commuter Population
In American higher education, commuter students comprise 84% of the undergraduate
student body population nationally and are enrolled in all segments of the higher education
system including two-year, four-year, public, and private institutions (National Center for
Educational Statistics [NCES], 2010). The commuter population represents a growing presence
on college campuses, yet commuters remain the overlooked majority among the student body.
The term commuter refers to any student who does not reside in institution-owned housing on
campus (Jacoby, 1989; Stewart & Rue, 1983). As defined, the commuter term is based on a
single criterion of residence off-campus, which has meant research and practice have

29

traditionally treated commuters as a homogenous group in comparison to their residential
counterparts (Astin, 1977, 1993; Pascarella et al., 1993). Reliance on the homogenous commuter
definition perpetuates the myths labeling all commuters as either “townies” assuming their place
of residence to be at home with parents, or assuming commuters are apathetic and uninterested in
the college experience (Jacoby, 2000b; Stewart & Rue, 1983).
This broad definition and limited concept of commuters is problematic because it
overlooks the diversity of needs and experiences within the commuter population (Jacoby &
Garland, 2004) and fails to account for variance in factors such as age, enrollment status, and
commuting distance (Kodama, 2002; Kuh et al., 2001). The commuter population is diverse and
can be described by characteristics including traditional and non-traditional age, part and fulltime attendance, native or transfer status, independence or dependence as related to parents,
employment status, and commuting distance to campus (Council for the Advancement of
Standards in Higher Education [CAS], 2012; Kodama, 2002; Kuh et al., 2001; Stewart & Rue,
1983). In higher education, commuter students represent an increase in the commuter proportion
of the student population and an expanding diversity among the student body. As such, further
understanding of the diversity within the commuter population is warranted and needed.
Challenges and Unique Needs of Commuters
The practice of commuting most notably influences the educational experience for
commuter students (Jacoby & Garland, 2004). Commuter students face challenges that place
demands on the time and energy they have for their educational experience. In addition to being
students, commuters often work to defray the costs of higher education and to provide support
for others as parents or caregivers (Kuh, 2003; Silverman, Aliabadi, & Stiles, 2008). Commuters
share concerns with transportation, including parking, traffic, vehicle maintenance, and cost

30

(Jacoby & Garland, 2004). Keeling (1999) described commuter students as “reinvented”
students, emphasizing the complexity of students’ lives managing multiple roles as employees,
parents, partners, and caregivers and underscored ‘student’ is no longer every students’ primary
identity (p. 4). The practice of commuting complicates the balancing act for students and
stresses the need for adequate support structures within the institution and beyond.
Demands of employment and multiple roles compete for a commuter student’s attention
and influence the amount of time, energy, and commitment given to academics. Competing
responsibilities mean commuters spend little time on campus overall and what time they do
spend is typically associated solely with class time (Krause, 2007). Compared to residential
peers, commuters are less involved in academic and social systems of an institution (Astin, 1977,
1993; Pascarella et al., 1993; Tinto, 1993), have less access to campus resources and services
(Krause, 2007), and fewer opportunities to develop relationships with faculty and peers because
they spend less time on campus overall (Jacoby & Garland, 2004). Given these obstacles,
commuters report a greater sense of marginality and a lessened sense of belonging to the
university (Jacoby & Garland, 2004; McLaughlin, 2011; Silverman et al., 2008). A lessened
sense of belonging is concerning because belonging is linked to positive college outcomes
including achievement, retention, and persistence (Hausmann et al., 2009; Rhee, 2008).
Despite the diversity of the commuter population, a common core of needs and concerns
was identified (Jacoby, 2000b). Key among them is the need to develop a sense of belonging to
the institution from addressing basic environmental needs such as lockers and lounges to
facilitating more intentional opportunities to engage with peers and faculty (Silverman et al.,
2008). Cultivating a sense of belonging helps commuters to “feel wanted” by the institution and
reduces feelings of marginality through the establishment of relationships and connection to

31

others (Jacoby, 1989, p. 6). Commuter students need assistance identifying and integrating
support systems into their educational experience (Jacoby, 2000a) as most existing support
systems are found in off-campus relationships with friends, family, and coworkers. Although
these support systems are important, often commuters must negotiate responsibilities, priorities,
and commitments related to academics and life roles with these same individuals. The realities
of support system complexities and belonging needs underscore the importance of the university
role in assisting commuters to establish relationships with staff and advisors who can aid in
promoting campus engagement and access to resources and services (Jacoby, 1989).
Need for Further Understanding of the Commuter Population
The known challenges and needs of commuters merit the attention of those in higher
education. If all students are to achieve maximum growth and learning in college, institutions
must expand their understanding of the commuter population and look to offer support
commensurate with the diversity within the commuter population. Krause (2007) argued
institutions must make efforts to ensure commuter voices are heard so their needs can be
“integrated into the structure of the institutional fabric at the classroom, academic department,
and broader institutional level” (p. 42). Dugan et al. (2008) caution against the “continued
utilization of programs and interventions designed for residential students with commuter student
populations, under the assumption that the effect on learning will be equivalent” (p. 283).
Yet, the dominance of the residential tradition continues to shape policies and practices in
higher education because research and practice have not kept pace with the changing commuter
population as it has diversified and grown. Continued reliance on a homogenous concept of
commuter students has allowed institutions to ignore the possibility within-group differences
may exist (Dugan et al., 2008; Kodama, 2002; Roe Clark, 2006) and has hampered efforts to

32

ensure all commuter students get the support needed to be successful in college. Further study of
the commuter population is warranted and needed if universities intend to offer every student
equitable access to the college experience regardless of place of residence (CAS, 2012).
As exemplified in the context of town-gown relations, an ongoing reliance on the
antiquated commuter concept has equated to an absence of information about off-campus
students and an unfamiliarity with how the college experience of off-campus students may be
similar or distinct from that of the traditional commuter population. As a result, efforts to
address town-gown tensions and to adequately support off-campus students are hampered
without empirical understanding of the off-campus student population. As such, the current
study focused on gaining insight and understanding of the off-campus student population and
their collegiate experience from a place of residence in the local community.
Summary of Commuter Literature
The commuter literature reviewed introduced off-campus students as central actors within
the town-gown context and illustrated their historical inclusion within the commuter population.
Further review of commuter scholarship explained the antiquated, homogenous definition of
commuter students in light of the growing presence of a diverse commuter population in higher
education. Challenges facing commuters included balancing multiple-life roles and employment
in college, while needs included gaining a sense of belonging and establishing support structures
to aid in campus involvement and integration. The limited understanding of the diverse
commuter population coupled with known challenges and needs of commuters warranted a call
for universities to support all students equally, regardless of place of residence. Further study of
off-campus students was warranted and drove the purpose of the current study, as the absence of
insight on off-campus students hampers efforts to address town-gown tensions and to adequately

33

support these students. Next, the literature review focuses on belonging literature to establish a
deeper understanding of the conceptual framework employed in the current study.
Conceptual Framework Literature
The final segment of the literature review focused on belonging literature in higher
education as the basis for the theoretical framework used to examine students’ behavioral
motivations in the local community. The review begins with an overview of the theories of need
to belong (Baumeister & Leary, 1995) and sense of community (McMillan & Chavis, 1986) as
they formed the blended theoretical foundation for the study. Then, each theory is discussed
with regards to the selection of variables used in the study. The section concludes with a review
of belonging literature in higher education and identifies the contribution of the current study.
A Blended Framework: Need to Belong and Sense of Community
The theoretical framework for the current study was built on two distinct, but related
theories, the theory of the need to belong (Baumeister & Leary, 1995) and the theory of sense of
community (McMillan & Chavis, 1986). Both theories focus on the human desire to belong to a
group or community and consider influences and outcomes of belonging. The theory of the need
to belong provided a construct to examine how the desire to belong served to motivate behavior
in students, while the theory of sense of community designated factors to consider for their
ability to influence belonging in a community. The following overview provides insight into the
theoretical underpinnings of each theory and discusses their use within the current study.
Theory of the need to belong. The primary conceptual anchor for the current study was
the theory of the need to belong as framed by Baumeister and Leary (1995) as a fundamental
human motivation. Specifically, the drive to satisfy the innate human need to belong is what
elicits affective responses and behaviors. The authors demonstrated the need to belong was a

34

powerful construct capable of producing emotional, behavioral, cognitive, and health outcomes
among humans. Utilized in higher education, the need to belong construct is conceptualized as a
sense of belonging construct and similarly noted for its ability to motivate and influence behavior
(Strayhorn, 2012). In college, sense of belonging refers to a student’s perceived social support,
feelings of mattering, and being connected to the campus community and others on campus
including students, faculty, and staff. Framed in this manner, sense of belonging is a cognitive
evaluation an individual makes with regards to their sense of identification in relation to a group
or community. As such, the desire to have a sense of belonging is a pervasive psychological
need and the drive to satisfy this need is what elicits affective responses and behaviors. In
education, satisfying the need to belong leads to positive, productive outcomes including
engagement, achievement, retention, persistence, optional functioning, and well-being
(Hausmann et al., 2009; Rhee, 2008; Strayhorn, 2012).
In the current study, the motivating capacity of the need to belong was of key importance.
As previously described, there are multiple positive outcomes known to be elicited from
achieving a sense of belonging. Thus, there was merit in exploring how belonging might
motivate off campus students’ behaviors. In the past, studies of college student belonging have
largely focused on belonging to the university or community entities within. The emphasis of the
current study on the town-gown context and the local community environment created a need for
a second theoretical component to shed light on the role of belonging to a community external to
the university. The theory of sense of community (McMillan & Chavis, 1986) and
corresponding model (Chavis & Wandersman, 1990) provided the needed theoretical lens for
examining the local community context.
Theory of sense of community. The theory of sense of community (McMillan &

35

Chavis, 1986) is closely related to the notion and concept of the need to belong as both pertain to
the interpersonal connections individuals desire and feel in relation to others. McMillan and
Chavis (1986) defined sense of community as “a feeling that members have of belonging, a
feeling that members matter to one another and to the group, and a shared faith that members’
needs will be met through their commitment to be together” (p. 9). Defined in this way,
McMillan and Chavis (1986) identified four essential elements in the generation of sense of
community and described them as:
• membership: the feeling of belonging or of sharing a sense of personal relatedness
• influence: a sense of mattering, of making a difference to a group and of the group mattering
to its members
• integration and fulfillment of needs: feeling that members’ needs will be met by the
resources received through their membership in the group
• shared emotional connection: the commitment and belief that members have shared and will
share history, common places, time together, and similar experiences
Taken together, the authors intended for the four elements to reflect the intimate, dynamic nature
of the development and maintenance of a sense of community.
Based on McMillan and Chavis' (1986) definition and concept of sense of community,
Chavis and Wandersman (1990) designed and tested a model of sense of community. The model
illustrated the catalytic effect sense of community has on local action (e.g., voluntary
participation in a neighborhood association) by affecting an individual’s perception of the
environment, social relations, and perceived control and empowerment within the environment.
The direction of causal influence of these variables was shown to be bidirectional and
transactional, as illustrated in the model when sense of community was found to be a cause and

36

result of local action. In the current study, the variables identified in the model as perception of
the environment and social relations are of key interest given their ability to influence and be
influenced by sense of community.
Application of the Conceptual Framework in the Current Study
The conceptual overview of the theories of the need to belong (Baumeister & Leary,
1995) and sense of community (McMillan & Chavis, 1986) has demonstrated the significance of
the desire for belonging in humans and shown the desire is sufficient to motivate behavior and
action in individuals. With this conceptual foundation established, the following section
delineates how both belonging theories were utilized in the design of the current study and the
selection of variables for study. The variables of interest corresponded with the research
questions below that guided the current study:
1. What is off-campus students’ sense of belonging to the local community? And, to what
extent do demographic characteristics and residential density contribute to the prediction
of off-campus students’ sense of belonging to the local community?
2. To what extent do environmental, social, and behavioral factors contribute to the
prediction of off-campus students’ sense of belonging to the local community?
3. What is the relationship between sense of belonging in off-campus students and their
involvement in positive behaviors (e.g., neighborhood involvement, responsible
neighbor) and negative behaviors (e.g., alcohol violations, litter, fights) within the local
community?
Examining sense of belonging. The first and second research questions focused on the
sense of belonging off-campus students have toward the local community. In higher education,
Strayhorn (2012) indicated sense of belonging relates to the satisfaction of a psychological need

37

to belong, such that students’ desire a “feeling or sensation of connectedness, the experience of
mattering or feeling cared about, accepted, respected, valued by, and important to the group (e.g.,
campus community) or others on campus (e.g., faculty, peers)” (p. 3). As such, variables
designed to measure off-campus students’ sense of belonging to the local community reflected
the psychological need to belong by focusing on perceptions of connection students perceived in
the local community.
Determining factors of influence on belonging. The second research question reflected
an intention to examine factors of potential influence on off-campus students’ sense of belonging
to the local community. Given the important linkages established between belonging and
positive outcomes of success (e.g., achievement, well-being), there was worth in determining
which factors influenced belonging and whether the factors contributed to or detracted from the
sense of belonging a student felt. This understanding of the influence of various factors could
enhance university efforts to cultivate belonging and promote success outcomes in students. In
selecting factors to consider for an ability to influence belonging, the model of sense of
community (Chavis & Wandersman, 1990) provided the following variables for consideration in
the current study:
• perception of environment refers to an individual’s judgements about the environment (e.g.
perceived qualities of and satisfaction with environment) and the “degree to which the
environment is positive or negative to the individual” (p. 57),
• one’s social relations refers to interactions among neighbors, such as informal visiting and
borrowing or lending tools, in which neighbors provide each other “emotional/personal,
instrumental, and informational support” (p. 58).
The third construct, namely the perceived control and empowerment items, were not used

38

because the focus of the study did not extend to understanding student agency or empowerment.
Examining sense of belonging and student behaviors. The final research question
centered on examining the relationship between sense of belonging and student participation in
positive and negative behaviors within the town-gown context. As previously explained, the
need to belong (Baumeister & Leary, 1995) is a fundamental human motivation and the drive to
satisfy this pervasive need to belong elicits affective behaviors and responses. The definition and
conceptual grounding of the need to belong provided the rationale for examining the relationship
between the need to belong and behavior in off-campus students. With regards to specific
behaviors of interest in the current study, the identification of positive and negative behaviors to
examine was informed by town-gown literature and practice.
Sense of Belonging Literature in Higher Education
Having articulated the conceptual framing of the current study and delineated the use of
the theoretical components in the selection of variables, the following section next reviews sense
of belonging literature within higher education. An overview of existing literature provides
insight into prior use of the theory of sense of belonging within higher education research and
captures the known findings and outcomes associated with the research.
Studies of underrepresented students. Within higher education, Hurtado and Carter
(1997) examined sense of belonging in relation to student persistence. Their study examined the
extent Latino students’ background characteristics and college experiences in the first and second
years contributed to sense of belonging in the third year. The findings of the study identified
significant positive relationships between sense of belonging and a variety of academic and
social activities, including peer tutoring and student organization involvement. Perceptions of a
hostile racial climate had a negative effect on the sense of belonging of the Latino student

39

participants. The authors found considerable variability in sense of belonging among students,
which was attributed to the differences in students’ perceptions of campus climates for diversity.
Building on the work of Hurtado and Carter (1997), Hausmann et al. (2009) examined the
effects of sense of belonging on persistence for Caucasian and African American first-year
students. The study revealed sense of belonging had direct effects on institutional commitment
and indirect effects on intentions to persist and actual persistence. The effects of belonging were
comparable for Caucasian and African American students for sense of belonging on institutional
commitment, intentions to persist, and actual persistence. An experimental intervention included
in the study was designed to promote sense of belonging in an intervention group with the
distribution of correspondence and gifts emphasizing the value of students as university
community members. Findings revealed the interventions positively influenced Caucasian
students but had no effect on African American students. Although specific strategies may differ
for various groups of students, the study demonstrated individual campuses could influence
sense of belonging and persistence through low-cost and easy to implement interventions.
In continued use of the sense of belonging construct with students of Color, Johnson et al.
(2007) expanded on Hurtado and Carter’s (1997) framework by studying African American,
Asian Pacific American, Hispanic/Latino, and Caucasian first-year students. The findings
revealed students of Color perceived a less strong sense of belonging than their Caucasian peers.
Students of all races and ethnicities who experienced a smooth transition to college perceived a
strong sense of belonging to campus. Findings related to perceptions of residence hall climate,
college transition, and campus racial climate suggested the shared responsibility between
students and their institutions for successful integration into the college community.
Studies of additional student demographics. Turning to another component of student

40

demographics, Ostrove and Long (2007) examined how social class background influenced sense
of belonging in college and had implications on college adjustment outcomes such as integration
and performance outcomes. Findings showed class background strongly related to students’
sense of belonging, which positively mediated social and academic adjustment, quality of college
experience, and academic performance. Findings suggested universities address collegiate
adjustment challenges of students from low social class backgrounds by cultivating belonging
with a welcoming, inclusive campus climate.
Kember, Lee, and Li (2001) studied the experience of part-time college students noting
most existing belonging research had focused on full-time students. The authors explored
whether part-time students had a sense of belonging and if so, the source of that belonging (i.e.
faculty, peers, academic department) and factors contributing to their belonging. Findings
suggested most students had a strong or developing sense of belonging and subsequently
identified class groups, faculty interactions, and departmental affiliation as the source of their
belonging. The authors offered strategies to foster belongingness, including increasing facultystudent interactions, focusing on initial academic experiences, and providing quality teaching.
Studies of the environment. Expanding beyond studies of student characteristics,
researchers have examined how factors of environment relate to sense of belonging. In
conjunction with their study examining sense of belonging with students of Color, Johnson et al.
(2007) also wanted to understand how perceptions of the residence hall environment related to
sense of belonging. The findings showed for students of all racial and ethnic backgrounds, sense
of belonging was strongly related to perceptions the residence hall climate was socially
supportive or tolerant of diversity. Residence halls appeared to be a compelling environment for
cultivating students’ sense of belonging because of the relationships formed and experiences

41

gained living in a residence hall. As such, the authors suggested the value of residence hall
settings came from the mutual responsibility students and staff share for fostering a living
environment culturally inclusive and supportive of all people.
Continuing a focus on environments, Freeman, Anderman, and Jensen (2007) examined
sense of belonging at class and campus levels for first-year students. At the class-level, they
examined associations between sense of belonging, class motivation, and characteristics of
effective teaching. At the campus-level, they examined how class-level belonging contributed to
campus-level belonging. Findings showed class-level belonging was an academic motivator for
first-year students and specific teaching practices, such as encouraging student participation and
instructor warmth, were strong contributors to belonging. At the campus-level, findings showed
class-level belonging did not contribute to overall campus-level belonging, but instead students’
sense of social acceptance was the largest contributor to campus belonging. In future research,
the authors called for the use of multiple data sources to better capture the dynamics of classlevel belonging processes and suggested better defining the relationship between class and
campus levels of belonging.
In the review of belonging literature in higher education, one commonality among all the
studies reviewed was an emphasis on sense of belonging and affiliation to the university campus
community and student relationships within this environment with peers, faculty, and staff. This
emphasis on a university centric focus was logical given the importance of understanding how
belonging influences the collegiate experience of students and ultimately influences outcomes of
interest within college, including achievement, retention, and persistence. However, the absence
of understanding how students’ feelings of belonging relate to communities and contexts external
to the university negates the reality that many students, particularly commuter and off-campus

42

students, blend experiences across university and community contexts as they navigate their
academic, residential, work, and social experiences. To better support off-campus students fully
in college, there was value in examining the sense of belonging and affiliation these students feel
to the local community where they reside and engage during their collegiate experience. Thus,
this study focused specifically on the sense of belonging students felt toward the local
community where they reside and examined how belonging influenced behaviors in the towngown context.
Summary of Conceptual Framework Literature
The purpose of the conceptual framework literature review was to explain the use of the
theories of need to belong and sense of community as the theoretical framework for the current
study and to explore existing belonging literature. The review began with an overview of the
concept and core elements of each theory to emphasize the strength of the desire for
belongingness as a fundamental human motivation. Next, the contributions of each theory were
delineated with regards to the selection of variables in alignment with the research questions
guiding the current study. Lastly, the review of belonging literature revealed prior use of sense
of belonging in higher education research with regards to underrepresented students, part-time
students, and environmental factors.
Contribution of the Current Study
The current study sought to advance higher education research and practice in multiple
ways. Within the context of town-gown practice and scholarship, the current study advanced
knowledge of off-campus students as central stakeholders in the town-gown relationship and
provided empirical understanding of this understudied portion of the student population. The
examination of off-campus students also provided insight into the sense of belonging and

43

affiliation these students feel toward the local community and offers practitioners perspective on
the influence of sense of belonging on students’ behavioral choices.
Within commuter scholarship, the current study responds to the call for additional study
and delineation of the diversity within the commuter population with an examination of offcampus students. In addition to advancing understanding of the demographic composition of the
off-campus student population, the current study provides practitioners with insight regarding the
similarities and differences between off-campus students and the broader commuter population.
Lastly, within sense of belonging literature, the focus of the current study on the local
community advances understanding of the role of sense of belonging in a context and
environment external to the university. The current study expands the use of the sense of
belonging framework with examination of factors capable of influencing belonging and the
examination of the role of belonging in eliciting positive and negative behaviors in students.
Summary of the Literature Review
The literature review focused on relevant literature along the three main themes of towngown relations, commuter students, and sense of belonging, to identify the context and need for
the current study. The review of town-gown literature introduced off-campus students and
detailed the concerns and challenges associated with these students in the town-gown context.
Further exploration of the dynamic, interdependent town-gown relationship provided
justification for further study of the off-campus student population. The review of commuter
literature revealed off-campus students have been traditionally included in the commuter
population, which was problematic given the antiquated, homogenous definition of commuters
does not reflect the known diversity within the commuter population. The absence of
information regarding off-campus students hampers efforts to adequately address town-gown

44

tensions related to student behavior issues. Lastly, the review of belonging literature offered
insight into the use of the theories of need to belong and sense of community as the conceptual
framework for the current study. The belonging construct provided opportunity to gauge
perceptions of belong with regards to the local community, to ascertain the influence of factors
on belonging, and to examine the relationship between belonging and behavior. Next, Chapter 3
will cover the methodology associated with the current study.

45

CHAPTER THREE
Methodology
This chapter elaborates on the methodology utilized in the current study. The chapter
includes a brief statement of the purpose of the study and its guiding research questions. A
discussion of the study design, the population and sample selection, data collection procedures,
instrumentation, and data analyses are also presented.
Purpose of Study
The purpose of the current study was to examine how sense of belonging served to
motivate and influence the behavior choices of off-campus students in the local community. The
desire for belonging is established as a fundamental human desire capable of motivating
behavior (Baumeister & Leary, 1995) and is utilized in higher education to demonstrate the
importance of belonging in helping students feel connected and engaged in college (Hurtado &
Carter, 1997; Ostrove & Long, 2007; Strayhorn, 2012). The current study aimed to further
student belonging research by using off-campus students as the population of interest and
focusing the inquiry on the nature of belongingness felt toward the local community as an
environment external, rather than internal, to the university. Furthermore, the current study
examined how sense of belonging served to motivate student behaviors in the local community.
The focus on student behaviors was chosen within the university-community context because
off-campus students are associated with town-gown tensions resulting from student behaviors in
the local community. The current study addressed the need for insight on off-campus students,
their affiliation to the local community, and their behavioral actions in a town-gown context.
Research Questions
With a research focus on examining off-campus students and their behaviors in the local

46

community, the following research questions guided the current study:
1. What is off-campus students’ sense of belonging to the local community? And, to what
extent do demographic characteristics and residential density contribute to the prediction of
off-campus students’ sense of belonging to the local community?
2. To what extent do environmental, social, and behavioral factors contribute to the prediction
of off-campus students’ sense of belonging to the local community?
3. What is the relationship between sense of belonging in off-campus students and their
involvement in positive behaviors (e.g., neighborhood involvement, responsible neighbor)
and negative behaviors (e.g., alcohol violations, litter, fights) within the local community?
Study Design
The current study was quantitative in nature and used a non-experimental survey design.
A quantitative approach aligned with the research focus on determining the relationship between
variables, namely sense of belonging, factors thought to influence belonging, and student
behaviors. A non-experimental design was appropriate given the intent on studying the variables
as they existed without manipulation by the researcher (Remler & Van Ryzin, 2011). The use of
a survey provided the opportunity to collect data from off-campus students in the spring semester
timeframe, in alignment with the notion college students often change residences between
academic years making a more longitudinal study of sense of belonging based on geographic
location difficult.
The first research question focused on evaluating the sense of belonging of off-campus
students, where sense of belonging was the criterion variable and demographic characteristics
and residential density categories were predictor variables. The second research question
examined the influence of environmental, social, and behavioral factors in predicting sense of

47

belonging in off-campus students. In this question, sense of belonging was again the criterion
variable, while the various factors were the predictor variables. The third research question
examined the relationship between off-campus students’ sense of belonging and their
participation in positive and negative behaviors. In this question, positive behaviors and
negative behaviors were the criterion variables, while sense of belonging, demographic
characteristics, density, environmental factors, and social factors were the predictor variables.
Institutional and Community Context
The current study examined off-campus students and the relationship between their sense
of belonging and behavioral choices within the local community. A review of both the
university and the local community contexts was needed to illustrate the town-gown relationship
and subsequent tensions resulting from student behavior concerns in the community. Examining
the institutional context further explained the focus on off-campus students as the population of
interest in the current study, while examining the local community context demonstrated the
significance of the presence of off-campus students in the local community.
Institutional Context
The current study was conducted at Michigan State University (MSU) which is a public,
research intensive, four-year institution located in East Lansing, Michigan. In the spring of
2016, MSU had a total student enrollment of 48,272 including 36,558 undergraduates, 10,191
graduate and professional students, and 1,533 non-degree seeking students (Michigan State
University, 2016). Of the total enrollment, 34,362 were in-state domestic students, 6,618 were
out-of-state domestic students, and 7,292 were international students. MSU has a first-year
residency requirement for undergraduates and boasts a large residential system housing 16,000
students, including both undergraduate and graduate students. Based on total student enrollment

48

and known quantity of residential students, an estimated 32,000 students reside off-campus in
non-institution owned properties, with the registrar’s office further estimating over 15,000 of
these off-campus students live in the local East Lansing community. The presence of a large
number of students residing in the local community provided an ideal setting to conduct the
current study focused on off-campus students and their behaviors within the town-gown context.
The institutional setting was purposely chosen as a residential university within a
“college town” community, as distinct from campus settings more rural or urban in nature
(Gumprecht, 2003, 2008). The “college town” setting was ideal for this study because the
university and local community were such that characteristics and actions of both entities had
sufficient influence on one another to demonstrate the intertwined nature of the town-gown
relationship. In particular, the college town served as a microcosm in which to observe how a
large number of students residing off-campus could influence the local community. Next, a
review of the local community context expands upon the town-gown relationship and the role of
students residing in a college town.
Local Community Context
The local community provided a context to observe the influence of the town-gown
relationship at play in the university and community interactions and the significance of their
influence on one another within a “college town” setting. The local community examined in this
study was the City of East Lansing, Michigan as the municipality where Michigan State
University is located. As such, the influence of the university on the local municipality could be
seen in several ways. MSU serves as the major employer within East Lansing with an estimated
11,000 employees in comparison to the next largest employer with 600 employees (City of East
Lansing, 2015a). In the 2010 U.S. Census, the local city population was 48,600 people with an

49

estimated 30,000 individuals between 18 and 24 years old and a disproportionally young median
age of residents at 21.6 years in contrast to 38.9 years state-wide (City of East Lansing, 2015b).
Most individuals in this age demographic were known to be college students who either lived oncampus in residence halls or in East Lansing within proximity to MSU.
Beyond population demographics, several additional features of East Lansing contributed
to its designation as a “college town”. Foremost, the City of East Lansing had a large rental
housing stock within the residential community, with 800 rental property owners or managers
responsible for over 1,600 licensed properties, ranging from individual houses to apartment
complexes (A. Irwin, personal communication, May 10, 2015). These licensed properties
constituted 9,500 rental units which equated to 27,000 occupants based on licensed occupancy
numbers. The predominance of rental properties within East Lansing reflected the need to
accommodate a transient college student population each leasing cycle and demonstrated the
sheer number of renters within the residential community. In addition, leasing agreements were
designed with students in mind as rental contracts were available as nine-month contracts aligned
with the academic year or as 12-month terms beginning and ending near the start of the fall
semester. The lease signing cycle was heavily influenced by the institution and its student
population with marketing and promotions taking place in October for leases beginning the
following August. The timeframe also coincided with MSU’s typical “Live On campaign”
designed to recruit students to sign up for on-campus housing for the upcoming academic year
(MSU, 2017). This advanced nature of lease signing was reflective of rental companies
competing with each other and the university for college student tenants and represented a timing
unique to university environments and in advance of traditional rental markets.

50

Sample Selection
The population of interest in the current study was undergraduate college students who
attended MSU and resided off-campus in the City of East Lansing. The population was
determined using residential criterion, specifically a current address in the municipal ZIP code of
48823. The term current address represented the address the student resides at while attending
MSU. Following IRB approval, the researcher submitted a data request to the MSU Office of the
Registrar (RO) for the sampling frame to be generated including the first and last name, MSU
email address, and street address for undergraduate students with a current address in the 48823
ZIP code in Spring 2016. The researcher received RO approval and an Excel file was generated
with 13,927 students meeting the sampling frame criteria. A review of the data revealed 461
cases of students who lived on-campus in the residence halls, but were erroneously included in
the data, likely because of an address error entered when each student updated his or her own
student directory information. Additionally, 34 cases were students who did not have complete
addresses, or the address was not in East Lansing. In total, 495 people were removed from the
population, resulting in an adjusted population of 13,432.
From the sampling frame, the sample was drawn using stratified random sampling. The
benefit of using random sampling comes from the use of probabilistic methods, which helps to
ensure statistical inferences can be drawn, a more representative sample can be achieved, and
sampling bias can be minimized (Remler & Van Ryzin, 2011). The use of sampling
stratification creates subgroups, or strata, within the sample and allows for comparison between
them. In the current study, residential density was of interest as a potential factor of influence in
off-campus students’ sense of belonging. Utilizing density as a factor allows for examination of
how proximity to others in a living environment influenced feelings of belonging. As such, the

51

stratum created reflected geographic areas of residence characterized as high, medium, and lowdensity areas of houses and apartments.
To create the strata designations for the study, the researcher worked with the City of
East Lansing’s geographical information systems (GIS) analyst to map the student address data
and discern patterns of high, medium, and low-density residency within East Lansing. Density
in the current study represented the number of off-campus students residing in a given land area.
The initial step was to map the student address data onto a City of East Lansing map, with each
student address represented by a dot on the map as seen in Map 1 in Appendix A. The broad
scale of Map 1 made individual address discernment difficult, but did illuminate initial areas of
density. To address this challenge and make density calculations possible, a grid was overlaid on
the East Lansing map with each grid unit representing the same size, 500 feet by 500 feet. The
grid overlay aligned closely with the size of a neighborhood block, while keeping consistent the
base unit of land across the entire City footprint. Map 2 in Appendix B depicts the grid overlay
on the East Lansing map with student data points included.
With the grid overlay in place, initial density calculations were computed and assessed
for clarity and meaning. The analysis revealed a striking difference between density levels
associated with houses and apartments, which resulted from the occupant capacity associated
with most apartments. Although house density was calculated on the block or grid level, not the
individual house level, most apartments were large enough to cover multiple grid units and had
sufficient occupants to appear to be high density nearly universally. The result was areas of
house density appeared less dense in comparison to apartments, while density calculations for
apartments were high at the individual grid level and no overall density calculation existed for
the apartment building. Rather than artificially segmenting apartments into smaller grid units,

52

the decision was made to recognize houses and apartments were distinct types of units and
density calculations were determined as such. For areas of houses, density calculations were
determined based on number of students within each grid unit of land. For apartments, density
calculations were determined based on number of students in relation to the licensed occupancy
rate of the apartment building or complex.
Once density calculations were made, the GIS software guided the identification of three
levels of density for each unit type (i.e., houses, apartments). The result was six density
categories representing high, medium, and low-density areas of both houses and apartments.
Map 3 in Appendix C is the map depiction of the density categories using color shading as
described in the map legend. Each student was placed in his or her corresponding density
category and the six categories became the strata levels for the current study. Map 4 in Appendix
D depicts both the shading and student data points. Once strata were identified, a random sample
for each stratum was drawn in a number proportional to the stratum’s size compared to the
population (Remler & Van Ryzin, 2011). These subsets of the strata were then pooled to form
the sample frame invited to participate in the current study. The final sample frame was n =
6,225, based on an estimated response rate of 6% for an online survey at MSU and a sampling
error of 5%. Table 2 shows the adjusted population and sample frame at each stratum level.
Table 2
Adjusted Population and Sample Frame Count at Each Stratum Level
Stratum
Population
Sample Frame
(N = 13,432)
(n = 6,225)
N
%
n
%
House High Density
1379
10.27
640
10.28
House Medium Density
1912
14.24
886
14.24
House Low Density
1925
14.33
892
14.33
Apartment High Density
4314
32.11
1999
32.11
Apartment Medium Density
2757
20.53
1278
20.53
Apartment Low Density
1145
8.52
530
8.52

53

Data Collection
Following the IRB approved study protocol, data for the current study were collected
using an online survey distributed to participants by email during a three-week timeframe in the
spring semester of 2016. The survey was designed and facilitated with Qualtrics, a survey
software available to the researcher through Michigan State University. Each student in the
study sample frame received an email at their official university email account with an invitation
to participate in the current study by completing the survey. Following the initial invitation to
participate, email reminders were sent to those students who had not yet participated in the study
or had yet to fully complete the survey instrument at the two-week interval. To encourage
participation, students who completed the survey could enter a drawing to win $100 toward their
rent. This incentive was paid for by the researcher and was coordinated with rental property
managers to ensure the money was used toward rent cost given the emphasis of the current study
on off-campus living. To protect participant identity, the Qualtrics software allowed individual
participation to be tracked in a manner that disassociated identifiable student information from
survey responses. Additionally, students who entered the incentive drawing did so by
completing a separate survey that collected their contact information for the drawing and was
distinct from the survey and their responses.
Instrument
The instrument used in the current study is comprised of questions drawn from three
distinct instruments; the Perceived Cohesiveness Scale (PCS) (Bollen & Hoyle, 1990), the Model
for Sense of Community scale (Chavez & Wandersman, 1990), and McLaughlin’s dissertation
instrument (2011). A combination of instruments was used to construct the current survey to
ensure there were questions to adequately address each of the study research questions. The

54

consent form and survey instrument are available for review in Appendix E and Appendix F,
respectively. Next, the instrument is reviewed and the selection of questions explained. For
reference, Appendix G has the list of questions associated with each construct.
The first research question examined the sense of belonging off-campus students have
toward their local community of residence. The sense of belonging questions were drawn from
the Perceived Cohesiveness Scale (PCS), which was used previously in higher education
literature (Hausmann, Schofield, & Woods, 2007; Hurtado & Carter, 1997; Maestas, Vaquera, &
Muñoz Zehr, 2007) as a measure of sense of belonging. The PCS measures an individual’s sense
of perceived cohesion, which is defined as “an individual’s sense of belonging to a particular
group and his or her feelings of morale associated with membership in the group” (Bollen &
Hoyle, 1990, p. 482). The PCS is comprised of two sets of questions; one related to sense of
belonging and the other one to feelings of morale. The current study used the set of three PCS
questions focused on sense of belonging to a group or community of interest. The three PCS
belonging items have a Cronbach’s alpha of 0.94 (Hurtado & Carter, 1997).
The second research question centered on understanding the influence community factors
had on the sense of belonging off-campus students felt toward the local community. Chavez and
Wandersman (1990) developed a model of sense of community (MSC) that demonstrated how
community factors influenced sense of community and in turn, sense of community acted as a
catalyst for community engagement. In the MSC, community factors were examined in three
themes: perceptions of the environment, social (neighbor) relations, and perceived control and
empowerment (Chavez & Wandersman, 1990). For the current study, questions addressing the
environmental and social relations themes were used. The environmental theme items came
from the MSC instrument and focused on assessment of block level problems and environment

55

concerns. The social theme included two sets of questions, both drawn from the McLaughlin
dissertation (2011), which like the current study, focused on student residents within a towngown context. The first set of social questions were based on the Perceived Neighborhood Scale
(Martinez, Black & Starr, 2002), and they assessed perceptions of sense of community, social
embeddedness, and satisfaction within a neighborhood. The seven neighborhood perception
items had a Cronbach's alpha of 0.87 (McLaughlin, 2011). The second set of social questions
focused on casual interactions and social relations with neighbors. The McLaughlin (2011)
questions were utilized, instead of similarly written MSC questions, because they offered
language reflective of a college environment. The neighbor relations items had a Cronbach’s
alpha of 0.91, (McLaughlin, 2011). Given social relations are likely to influence participation in
normative behaviors, both good and bad, the behaviors construct was included in the current
study as a factor for consideration as part of the social theme. The development of the behavior
construct is described in detail below.
The third research question centered on understanding the relationship between a
student’s sense of belonging and their participation in a range of behaviors in the local
community. The researcher developed an initial list of behaviors for the survey based on her
professional experience working in town-gown relations, in conjunction with a review of two
annual assessment surveys from the International Town-Gown Association (2013b, 2014), which
discussed student behavior challenges and other town-gown topics. The researcher then
consulted with two town-gown professionals, one employed by the university and one employed
by the city, to review the generated list and make suggested modifications. The final product
was a list of 18 behaviors representative of desired positive behaviors (e.g., meet neighbors,
participate in neighborhood events) and undesired negative behaviors (e.g., littering, hosting

56

parties, noise generation) central to town-gown tensions associated with off-campus students. In
responding to the behavior questions, participants were asked to indicate the frequency by which
they participated or engaged in each behavior.
Data Analysis
Survey data were analyzed using IMBM SPSS Statistics, Version 24 software. Planned
data analyses included the use of descriptive statistics and regression analyses. Descriptive
statistics were computed for all demographic variables (e.g., age, sex, race, year in school).
Multiple regression was used for each of the three research questions. Multiple regression
analysis was chosen as a dependable and widely-used technique in quantitative analysis in
educational research. Regression was an appropriate analysis method for each research question
because of the focus on predicting the influence of predictor variables on criterion variables
(Remler & Van Ryzin, 2011). The first research question examined how student demographic
factors and density influenced off-campus students’ sense of belonging to the local community.
The second research question concerned predicting the influence of environmental, social, and
behavioral factors on off-campus students’ sense of belonging. The third research question
related to predicting how sense of belonging influenced positive and negative student behavior
participation, while also taking into consideration the influence of demographics, density,
environmental factors, and social factors.
Limitations of the Study
There are strengths and weaknesses with every study. The crucial component in all
research endeavors is to identify these limitations and to understand their implications for the
study (Remler &Van Ryzin, 2011). The current study was exploratory in the sense off-campus
students had been largely overlooked in empirical research and sense of belonging had not been

57

utilized to examine town-gown relations and the off-campus context. That said, there are a few
limitations to note in the current study.
The design of the study captured the experience of off-campus students in the town-gown
context specific to Michigan State University and the City of East Lansing. As a single sample
collected in a specific context, the findings of the current study are unique to the sample and not
intended to generalize to the larger MSU student population and may differ across different
town-gown contexts. The chi-square goodness-of-fit analysis revealed a limitation in the data as
the sample and the population differed statistically indicating the sample was not representative
of the larger population. Overall, the sample studied had more females and more White students
than the population of off-campus students. Similarly, a racial imbalance existed in the studied
sample (see Table 3) with more White students participating in the study than students of Color.
Full details of the chi-square analysis are provided in Chapter 4 in conjunction with the overview
of the study sample. Despite the limitation of generalizability, the current study offers insight
about sense of belonging and off-campus students that town-gown administrators can utilize to
infer meaning and practical applicability of these findings to their given context.
The population from which the sample was drawn also represented a limitation because
the population was an incomplete count of off-campus students due to administrative restrictions.
Following IRB approval, the source of the population data was the University Registrar. As a
practice, the registrar’s office does not include information in data releases for students with
restrictions in place on their account, such as voluntarily restriction of their address. Due to this
administrative policy alone, the registrar’s office withheld 1,832 students from the off-campus
student population being considered in the study (J. Murphy, personal communication, February
16, 2016). Additionally, the registrar’s office acknowledged, without providing a quantifiable

58

number, students are not required to submit a local address in the student information system,
although the majority do. Together, these limitations reduced the starting population from which
the identified sample was drawn.
The self-reported nature of the questions in the study survey was also a limitation.
Humans tend to report experiences in surveys in a way that show them in a better light, thus not
guaranteeing the accuracy of the results (Remler & Van Ryzin, 2011). The current study also
relied on participants to answer questions about a range of behaviors, including punishable or
risky behaviors such as drinking or fighting. The survey did not have a connection to municipal,
judicial, police, or university conduct records, but there could have been concern from
participants of a possible linkage or potential repercussions from participating. To address this
concern, the survey consent form indicated participant identity was disassociated from their
responses, thus making it impossible to connect participants to responses or to pursue any
judicial repercussions.
Conclusion
This chapter included a review of the methodology associated with the current study. A
town-gown context provided a rich setting to study off-campus students’ feelings of belonging to
the local community where they reside during college. The current survey instrument combined
questions about perceived cohesion, factors influencing community living, and student behaviors
to explore the sense of belonging off-campus students felt toward the local community. Planned
data analyses included the use of descriptive statistics and regression analyses. Next, the survey
results and findings are presented in Chapter 4.

59

CHAPTER FOUR
Research Findings
The purpose of this study was to examine the sense of belonging off-campus students felt
toward the local community they reside in. Additionally, the study sought to examine the
potential influence of environmental, social, and behavioral factors on sense of belonging, and if,
in turn, sense of belonging could influence participation in positive and negative behaviors in the
town-gown context. The chapter presentation begins with details of preliminary data screening,
sample demographic composition, and construct development. Then, multiple regression analysis
is discussed and regression assumptions are tested. Finally, regression analyses for each research
question are presented and discussed. The chapter concludes with a summary statement.
Preliminary Data Screening
The initial dataset for analysis contained 645 cases, in correspondence to the number of
completed surveys received in the Sense of Belonging survey. Before primary data analysis
began, the dataset was checked for errors and missing values. Ensuring the accuracy of data
prior to primary analysis was essential for credible analysis. First, frequencies and descriptive
statistics were calculated for each variable in the dataset. This review identified errors in data
entry and coding in the dataset. Data entry errors were resolved by comparing the raw data
output from Qualtrics with the SPSS dataset and making necessary adjustments. No coding
errors were identified through the examination of frequencies.
Second, univariate statistics were examined for each variable and case to determine the
frequency of missing values. Analysis showed none of the 53 individual variables were missing
more than 1% of responses, so all variables were retained. The review of participant cases
showed 36 cases had missing values with 31 cases missing one response (1.7%), three missing

60

two responses (3.4%), one missing four responses (7.4%), and one missing 20 responses
(38.3%). The two cases with greater than 5% of missing values were deleted from the dataset
(Tabachnick & Fidell, 2007), leaving a reduced sample size of 643. The pattern of missing
values was assessed with a Missing Values Analysis (MVA) in SPSS, which showed data were
missing completely at random (MCAR) with a non-significant Little’s MCAR test, X2 (1246.41,
n = 643) = 1223.19, p = .314. Given the random pattern and low frequency of missing values,
the decision was made to use list-wise deletion for missing values during analysis (Tabachnick &
Fidell, 2007). List-wise deletion removes a case if a missing value existed in the variables being
considered, so only complete cases are analyzed (IBM, n.d.). List-wise removal of missing
values reduces the risk of a Type-1 error.
Sample Demographics
The population (N) in the current study was comprised of 13,432 undergraduate students
who lived in the target ZIP code of 48823 in the City of East Lansing in spring semester 2016.
Using a stratified random sampling strategy, a sample frame of 6,225 potential respondents was
selected for the current study. Full details of the sampling strategy were described in Chapter 3.
A total of 645 respondents submitted completed surveys for a response rate of 10.4%. In
preliminary screening, two cases were dropped due to excessive missing values, resulting in a
reduced sample of 643. The sample was further reduced during normality screening of variables
with the list-wise deletion of 34 cases and the removal of 20 outlier values. Discussion of the 54
removed cases is provided in a later section detailing testing of normality assumptions. The final
reduced sample size was 589, which represented retention of 92% of the original sample.
With the emphasis on density stratification in sample selection, the strata membership of
participants was analyzed with a chi-square goodness-of-fit test. The chi-square goodness-of-fit

61

test compares the proportion of sample participants (n = 589) with values from a comparison
population, namely, the study population (N=13,432) and corresponding stratification (Ott &
Longnecker, 2001). The chi-square goodness-of-fit test indicated there was no significant
difference in the proportion of sample participants in each of six stratums as compared to the
observed values of each stratum in the study population, X2 (5, n = 589) = 4.95, p = .422.
Participant demographics were self-reported in the survey for characteristics including:
race, sex, age in years, year in school, number of roommates, and semesters lived on-campus and
off-campus. For the current study, race data were collapsed and individuals were classified as
White, students of Color, or International because of the small number of minority and
international participants. The students of Color category was inclusive of African Americans,
American Indians, Asians, Hispanics, Hawaiian/Native Pacific Islanders, and Multiracial
individuals. The international category reflected the university use of an international choice in
the race category, distinct from the minority groups referenced above.
Inspection of frequencies in the demographic data (n = 589) revealed most of the
participants were female (63.8%) and White (84.0%). With a mean of 21 years (36.7%), most
participants were in their third year (38.0%) or fourth year (38.7%) of college. The most
frequent number of roommates was three (34.5%). Most participants reported having lived oncampus two semesters (46.9%) and living off-campus two semesters (35.8%). Table 3 provides
participant demographic characteristics for the original sample (n =643) and the reduced final
sample (n=589).

62

Table 3
Participant Demographic Characteristics for Sample and Reduced Sample
Demographic variable

Sample (n = 643)
n
%

Reduced Samplea (n = 589)
n
%

Sex
Female
Male

413
230

64.2
35.8

376
213

63.8
36.2

Race
White
Students of Color
International

535
75
33

83.2
11.7
5.1

495
65
29

84.0
11.0
4.9

Age in years
18
19
20
21
22
23
24
25 or greater

3
56
167
229
129
37
8
14

0.5
8.7
26.0
35.6
20.1
5.8
1.2
2.5

3
50
150
216
121
36
7
6

0.5
8.5
25.5
36.7
20.5
6.1
1.2
1.0

Year in School
First year
Second year
Third year
Fourth year
Fifth year
Other

3
108
243
244
37
8

0.5
16.8
37.9
38.0
5.8
1.2

3
92
224
228
35
7

0.5
15.6
38.0
38.7
5.9
1.2

Semesters Lived On-Campus
None
One
Two
Three
Four

99
51
300
16
177

15.4
7.9
46.7
2.5
27.5

88
45
276
15
165

14.9
7.6
46.9
2.5
28.0

Semesters Lived Off-Campus
One
Two
Three
Four
Five

14
231
23
204
171

2.2
35.9
3.6
31.7
26.6

12
211
21
186
159

2.0
35.8
3.6
31.6
27.0

63

Table 3 (cont’d)
Sample (n = 643)
n
%
Number of roommates
None/Lives alone
One
Two
Three
Four
Five
Six
Seven
Eight
Ten or greater

39
156
90
217
57
28
9
8
3
35

6.1
24.3
14.0
33.7
8.9
4.4
1.4
1.2
0.5
5.5

Reduced Samplea (n = 589)
n
%
36
146
87
203
56
25
9
7
3
17

6.1
24.8
14.8
34.5
9.5
4.2
1.5
1.2
0.5
2.9

Note. a the reduced sample size was the result of list-wise deletion of missing values and the
removal of outliers during normality screening of variables.
To assess if the sample participants were representative of a comparison off-campus
student population, a series of chi square goodness-of-fit tests were performed. Participant
demographics for sex, race, and year in school were assessed with a comparison population (N =
16,312) of undergraduate students residing in the East Lansing ZIP code (48823) during the Fall
2015 semester (S. Webster, personal communication, January 18, 2017). For each demographic
examined, the chi-square goodness-of-fit tests indicated a significant difference between the
sample and comparison population. For sex, the sample had more females (64%) and fewer
males (36%) than the population proportion of females (48%) and males (52%), X2 (1, n = 589) =
59.19, p < .001. For ethnicity, the sample had more White students (84%), fewer students of
Color (11%), and fewer international students (4.9%) than the population proportion of White
students (70%), students of Color (13%), and international students (17%), X2 (2, n = 589) =
67.24, p < .001. For year in school, the sample has fewer first year (0.5%) and second year
(17%) students and more third year (41%) and fourth year (42%) students than the population
proportion of first year (4%), second year (21%), third year (34%) and forth year (40%) students,

64

X2 (3, n = 546) = 31.20, p < .001. The chi-square analysis for year in school had fewer cases
than the other chi-square analyses because the comparison population was limited to first
through fourth year students, whereas the study included fifth year respondents. The variable was
recoded to remove fifth year students for chi-square analysis, resulting in n = 546. These
differences indicated the sample was not representative of the comparison population. As such,
study findings should not be generalized to the off-campus student population broadly and
interpretations should be done with awareness of the limitation. Nevertheless, the current study
provided initial insight into the sense of belonging of off-campus students in the MSU-East
Lansing context.
Construct Development
Six constructs were explored through hierarchical regression analysis in the current study.
As noted in Chapter 3, four of the six constructs came from theoretical and empirical literature
on sense of belonging (Bollen & Hoyle, 1990), sense of community (Chavez & Wandersman,
1990; McLaughlin, 2011), and community ties (McLaughlin, 2011). The fifth (PosBehav) and
sixth (NegBehav) were researcher created constructs generated from an existing town-gown
association survey (ITGA 2013b; ITGA 2014), then reviewed and further edited by local towngown professionals. The basis in an existing survey and the subsequent review by professionals
with content expertise helped to address concerns with content validity of the new measure
(Remler & Van Ryzin, 2011). Each construct scale was a combination of scores from Likertstyle questions on the study’s survey instrument. The construct names and abbreviations are
reported in Table 4 and items used to develop each construct are provided in Appendix G.

65

Table 4
Cronbach’s Alphas for Constructs Used in Multiple Regression Analyses
Construct
Construct abbreviation
Cronbach’s alpha
Sense of Belonging Total
SoBTotal
.91
Environmental Factors
Factors
.79
Neighborhood Perceptions
Perceptions
.87
Neighborhood Relations
Relations
.91
Positive Town-Gown Behaviors
PosBehav
.61
Negative Town-Gown Behaviors
NegBehav
.74

Construct Reliability and Validity
The reliability of a construct concerns the consistency of the measure over time and over
a variety of conditions. For constructs composed of multiple items, reliability is a measure of the
internal consistency between the various items. A widely-used measure of internal consistency
reliability is Cronbach’s coefficient alpha (a) (DeVellis, 1991). The range for Cronbach’s a
values is 0 to 1, with values of .70 considered sufficient and higher values seen as respectable,
while more modest values of .50 to .60 as acceptable for the beginning stages of research
(DeVellis, 1991). For each construct used in the current study, a Cronbach’s a was calculated
and reported in Table 4 above. The calculated Cronbach’s a for the constructs ranged from .61
to .91, all within the range of sufficient and respectable values for internal consistency reliability.
The development of each construct is described next. The survey questions utilized in
the development of the construct are referenced along with their corresponding variable name in
parenthesis (e.g., Sex). Table 5 has descriptive statistics (e.g., mean, SD) for each construct.
Background Characteristics
Background characteristics consisted of seven variables including: age in years (Age),
sex (Sex), the dummy coded race variable (Race1, White; Race2, students of Color; Race3,
International), year in school (YrSchool), number of semesters lived on-campus (LiveOn) and

66

off-campus (LiveOff), and number of roommates (RoomLog). Six density variables (AptHigh,
AptMed, AptLow, ResHigh, ResMed, ResLow) were used as background characteristics because
density was an important study component and a central part of the stratified sampling strategy.
Sense of Belonging Construct
Sense of belonging relates to the human desire to feel connection and belongingness with
others in the community (Hurtado & Carter, 1997; Strayhorn, 2012). In the current study,
respondents answered three questions about their sense of belonging to the local, off-campus
community. The three belonging questions were: “I feel that I am a member of the East Lansing
community” (ELmember), “I see myself as part of the East Lansing community” (PartofEL), and
“I feel a sense of belonging to the City of East Lansing” (SoBtoEL). The Likert-scale scores (1=
Strongly Disagree, 2= Disagree, 3= Neutral, 4= Agree, 5= Strongly Agree) of the three questions
were summed to create the Sense of Belonging construct (SoBT). These questions and the
derived SoBT construct were modeled after work of Bollen and Hoyle (1990) and others
(Hausmann et al., 2007; Hurtado & Carter, 1997). With a Cronbach’s a of .91, the SoBT
construct had strong internal reliability as a measure of sense of belonging and was consistent
with the previous works identified.
Environmental Factors Construct
The environmental factors construct stemmed from the work of Chavez and Wandersman
(1990) that examined how perceptions of the environment influenced sense of community. The
questions related to potential problems present in an environment at the neighborhood block
level and asked respondents to indicate the significance of each problem on a Likert-scale (1 =
Not a problem, 2 = Minor problem, 3 = Moderate problem, 4 = Major problem). The 11
environment questions used to derive the Environmental Factors construct (Factors) were:

67

“Conditions of houses” (Houses), “Traffic” (Traffic), “Street parking” (Parking), “Crime”
(Crime), “Litter” (Litter), “Noisy Neighbors” (Noise), “Unkempt lawns” (Lawns), “Garbage
collection” (Garbage), “Recycling collection” (Recycling), “Police service” (Police), and
“Parking services-enforcement” (PACE). Low scores on the Factors scale represented a low
perception of community problems and high scores represented a high perception of community
problems. Cronbach’s a for the Factors construct was .79, which indicated a respectable internal
reliability as a measure of environmental factors in assessing community belonging.
Neighborhood Perceptions Construct
Like the environmental factors construct, the neighborhood perceptions construct
examined how perceptions influence feelings of community belonging, with an emphasis on
perceptions of neighborhood relations. Drawn from the dissertation work of McLaughlin (2011),
six questions made up the neighborhood perceptions construct (Perceptions) and utilized a Likert
scale (1 = Strongly Disagree, 2 = Disagree, 3 = Neutral, 4 = Agree, 5 = Strongly Agree). The
construct items were: “There are people I can rely on among my neighbors” (RelyOn), “People
trust each other in my neighborhood” (TrustOthers), “I feel I belong in my neighborhood”
(BelongTo), “I care about what my neighbors think of my actions” (Care), “I feel close to some
of my neighbors” (CloseTo), “People in my neighborhood are usually warm and friendly”
(Friendly), and “We look out for one another in this neighborhood” (LookOut). The Perceptions
construct had a Cronbach’s a of .87, which indicated good reliability in measuring neighborhood
perceptions and was consistent with prior work with the same construct (McLaughlin, 2011).
Neighborhood Relations Construct
The neighborhood relations construct focused on neighborhood actions and social
interactions among neighbors. The neighborhood relations construct (Relations) utilized two sets

68

of questions to capture this construct (McLaughlin, 2011). The first set used a Likert scale (1 =
Very Unlikely, 2 = Unlikely, 3 = Not Sure, 4 = Likely, 5 = Very Likely) and each question began
“How likely is it that you would…” and included: “Borrow something like a tool, if they had
one” (BorrowTool), “Borrow something like money” (BorrowMoney), “Ask you to borrow
something, like a tool, if you had one (LoanTool), “Ask you to borrow something, like money
(LoanMoney), “Get help from a neighbor” (HelpFrom), and “Help a neighbor” (HelpTo).
The second set of questions used a Likert scale (1 = Very Seldom/Never, 2 = Seldom
(once every 3 months), 3 = Sometimes (once per month), 4 = Often (once per week), 5 = Very
Often (daily)) and each question began “How often do you…” including: “Greet your neighbors
when you see them” (Greet), “Casually visit with neighbors, either going over to their place or
their coming over to yours” (CasualVisit), “Participate in neighborhood activities (e.g., cookouts,
parties)” (NeighActiv), “Talk to neighbors who are students” (TalkToStu), “Talk to neighbors
who are non-students” (TalkToNonStu), and lastly, “People in this neighborhood do favors for
each other” (FavorsFor). Taken together, the 12-item Relations construct had a Cronbach’s a of
.91, which indicated a good reliability as a measure for neighborhood relations and was
consistent with prior work with the same construct (McLaughlin, 2011).
Positive and Negative Behaviors Constructs
The positive behaviors construct (PosBehav) and negative behaviors construct
(NegBehav) both gauged the frequency of student participation in behaviors common in the
town-gown context. The construct items were derived from an International Town-Gown
Association (2013b, 2014) assessment and were vetted by university and municipality
professionals with town-gown expertise. The 12 items represented a set of five positive or
desired behaviors and a set of seven negative or more problematic behaviors within the town-

69

gown context. To better understand the role of behaviors in the study, a positive behaviors
construct and a negative behaviors construct were created. Both constructs were derived from
items measured on a Likert scale (1 = Very Seldom/Never, 2 = Seldom (once every 3 months), 3
= Sometimes (once per month), 4 = Often (once per week), 5 = Very Often (daily)). Each
question began with the prompt “how often do you…” and asked respondents to indicate
frequency of participation in a behavior.
Positive behavior construct. The positive behaviors construct (PosBehav) included five
behavior items: “Do yard maintenance, like mowing, raking, or shoveling” (YardMaint), “Use
trash and recycling services at your residence” (TrashRecy), “Pick up litter or trash in your yard”
(Litter), “Participate in community-wide activities (e.g., Farmer’s Market, live music)”
(CommActiv), and “Use city resources or services (e.g., public library, community center, parks,
recreation facilities)” (CityResource). The positive behaviors construct had a Cronbach’s a of
.61, which is modest, but acceptable in early stages of research (DeVillis, 1991) as was the case
with this new construct for evaluating participation in positive behaviors in a town-gown context.
Negative behavior construct. The negative behaviors construct (NegBehav) included
seven items: “Attend a party in your neighborhood” (PartyAttend), “Host a party at your
residence” (PartyHost), “Consume alcohol” (Alcohol), “Receive an alcohol violation from
police” (AlcoholViol), “Have loud noise coming from your residence (e.g. music)” (NoiseFrom),
“Receive a noise violation ticket at your residence” (NoiseViol), and “Get into a fight with
people in your neighborhood” (Fight). The negative behaviors construct had a Cronbach’s a of
.74, which indicated a suitable reliability as a measure of participation in negative behaviors in
the town-gown context.

70

Table 5
Descriptive Statistics for Constructs (n = 589)
Construct (Variable name)
Sense of Belonging Total (SoBT)
Environmental Factors (Factors)
Neighborhood Perceptions (Perceptions)
Neighborhood Relations (Relations)
Positive Behaviors (PosBehav)
Negative Behaviors (NegBehav)

Mean
11.28
21.45
22.79
27.97
12.79
13.61

SD
2.61
5.88
5.33
9.90
3.51
4.03

Minimum
3
11
7
12
5
7

Maximum
15
41
35
60
22
28

Regression Analysis
Multiple regression is a set of statistical techniques used to explore the relationship
between one criterion variable and one or more predictor variables. In hierarchical multiple
regression, predictor variables are entered into a model in an order specified by the researcher
and rooted in theoretical grounds. Predictors are entered in steps, with each variable (or set of
variables) being evaluated for its contribution to the model at its own time of entry, after
accounting for the contributions of previous variables (Tabachnick & Fidell, 2007). In the
current study, the order of predictor variable entry is discussed as a component of the analysis for
each research question.
For each research question analyzed, the results of the corresponding hierarchical
regression analysis are presented in table format in the appendices. Each table reports the
standardized regression coefficients (b), R, R squared (R2), adjusted R2 (Adj. R2), change in R2
(∆R2) and change in F (∆F) associated with each model (Tabachnick & Fidell, 2007).
Standardized regression coefficients (b) represent the contribution of each variable to the
prediction of the dependent variable, given the other variables at the same stage in the model.
The standardized coefficients were chosen, over unstandardized coefficients, so variables could
be compared to one another on the same scale. R2 is the proportion of variance in the dependent

71

variable explained by the regression model. Adjusted R2 corrects for interactions between
variables and provides the adjusted R2 value that would result from a new sample being run on
the same set of data. In hierarchical regression, an R2 value is calculated for each step in the
model, but as steps are added, the R2 value includes the proportion of variance provided by the
current step and earlier steps. The incremental change of R2 at each step is represented by the
change in R2 (∆R2) value. The change in F (∆F) determines the significance of the ∆R2 value at
each step. Statistical significance was set at .05 for all analyses, with a 95% confidence interval.
Evaluation of Regression Assumptions
Prior to performing analysis to test the research questions, 34 cases were removed via
list-wise deletion, resulting in a starting sample of 609 for normality analysis. Next, assumptions
of normality were evaluated for interval and ratio variables with examination of means, standard
deviations, minimum and maximum values, and skewness (Table 6). Normal distributions were
seen across all variables except for Age and Roommates, which had severe positive skewness
values of 9.33 and 5.35, respectively. Univariate outliers were found in Age and Roommates by
the identification of cases with z-scores more than three standard deviations from the mean, (p <
.001, two-tailed test) (Tabachnick & Fidell, 2007). For Age, six cases with outlier values were
removed, resulting in an improved skew value of .52. For Roommates, 14 outlier cases were
removed and a logarithm transformation was performed, resulting in an improved skew value of
.19 for RoomLog. In total, 20 outliers were removed leaving a final reduced sample of 589. The
reduced sample size was adequate for analysis with 16 predictor variables (Tabachnick & Fidell,
2007). Descriptive statistics were recalculated in Table 7 for all variables using n = 589.

72

Table 6
Descriptive Statistics for Interval and Ratio Variables After List-Wise Deletion (n = 609)
Variable Name
Age
Year in School
Live On Semesters
Live Off Semesters
Roommates
Sense of Belonging
Environmental Factors
Neighbor Perceptions
Neighbor Relations
Positive Behaviors
Negative Behaviors

Mean

Standard Deviation

Skew

Minimum

Maximum

21.09
3.36
2.18
3.45
3.75
11.29
21.51
22.89
28.10
8.96
13.60

2.13
.88
1.33
1.28
6.80
2.58
5.99
5.35
9.86
3.19
4.02

9.33
.15
- .04
- .17
5.35
- .86
.52
- .11
.42
.41
.77

18
1
0
1
0
3
11
7
12
4
7

57
6
4
5
54
15
44
35
60
18
28

Table 7
Descriptive Statistics for Interval and Ratio Variables in Final Reduced Sample (n = 589)
Variable
Mean
Standard Deviation
Skew Minimum Maximum
Age
20.96
1.18
.52
18
25
Year in School
3.38
.88
.10
1
6
Live On (semesters)
2.21
1.33
- .06
0
4
Live Off (semesters)
3.46
1.28
- .18
1
5
Roommates (Logarithm
.52
.23
.19
0
1.40
Transform)
Sense of Belonging
11.28
2.61
- .85
3
15
Environmental Factors
21.45
5.88
.45
11
41
Neighbor Perceptions
22.79
5.33
- .09
7
35
Neighbor Relations
27.97
9.90
.44
12
60
Positive Behaviors
12.79
3.51
.28
5
22
Negative Behaviors
13.61
4.03
.76
7
28
The assumption of singularity was met because the predictor variables, both individually
and as constructs, were not combinations of other predictor variables. The assumption of
multicollinearity was met because predictor variables were not highly correlated and collinearity
statistics (i.e., Tolerance and VIF) were within accepted limits (Ott & Longnecker, 2001). See
Appendix H for the correlation matrix of predictor and criterion variables and see Appendix I for
collinearity statistics associated with each final regression model. Univariate outliers were
73

removed as described above and no multivariate outliers were found based on Mahalanobis
distance scores. Residual and scatter plots indicated assumptions of normality, linearity, and
heteroscedasticity were satisfied (Tabachnick & Fidell, 2007).
Sample Variance Tests for Sense of Belonging
Before beginning analysis for the research questions, sample variance tests were done to
ascertain if group differences existed in sense of belonging (SoBT) as a function of demographic
characteristics. Predictor variables with two distinct groups were analyzed with an independentsamples t-test, while ordinal variables with more than two groups were evaluated using one-way
analysis of variance (ANOVA) tests. Sex (male, female) was evaluated with a T-test. The
remainder of demographic variables evaluated with ANOVAs included: race, year in school,
semesters lived on campus, semesters lived off campus, and the density variable. Results of the
sample variance tests are presented in table format in Appendix J.
The independent-samples t-test was conducted to compare the sense of belonging scores
for males and females. There was a significant difference in mean scores for males (M = 11.00,
SD = 2.65) and females (M = 11.44, SD = 2.58; t (587) = 2.01, p = .045, two-tailed). The
magnitude of the differences in means (mean difference = .45, 95% CI: .01 to .89) was very
small (eta squared = .006). This finding indicated males had lower sense of belonging to the
local community than females.
For the ANOVAs conducted, there was no significant difference noted in scores for race,
year in school, semester lived off-campus, and density. The ANOVA for living on campus
(LiveOn) revealed a significant difference at the p < .05 level in the sense of belonging (SoBT)
scores: F (4, 584) = 3.40, p = .01. The magnitude of difference on mean scores between groups
was small (eta squared = .02). Post-hoc comparisons using Tukey HSD indicated the mean score

74

for No Semesters (M = 10.51, SD 2.69) differed from the means for One Semester (M = 12.07,
SD 2.21) and for Two Semesters (M = 11.39, SD 2.53). This finding indicated students who had
never lived on campus had a lower sense of belonging than students who had lived on campus
one semester or two semesters.
The following sections report the results of the three research questions in the current
study. Each question is restated and associated regression analyses are presented in a table and
interpretations of findings are provided. Final regression models for each question, including
standardized and unstandardized coefficients and collinearity are provided in Appendix I.
Research Question One
RQ 1: To what extent do demographic characteristics and residential density contribute to the
prediction of off-campus students’ sense of belonging to the local community?
The first research question sought to examine what effect background characteristics and
residential density had in predicting the sense of belonging of off campus students to the local
community. A two-step hierarchical regression was conducted with Sense of Belonging (SoBT)
as the dependent variable. Demographic characteristics including age (Age), sex (Sex), race
(Race2, students of Color; Race3, International), year in school (YrSchool), number of semesters
lived on campus (LiveOn), number of semesters lived off campus (LiveOff), and number of
roommates (logarithm transformation) (RoomLog) were entered as predictor variables in step 1,
with dummy coded Race1 (White students) left out as a constant. Step 2 included the addition of
residential density categories consisting of five dummy variables (AptMed, AptLow, ResHigh,
ResMed, ResLow) with the sixth dummy variable (AptHigh) left out as the constant.
At the end of step 1 and step 2, neither model showed a statistically significant R2
relationship between background characteristics or residential density in predicting sense of

75

belonging. None of the predictor variables were statistically significant predictors of sense of
belonging. The variance at the end of model 2 was .036 indicating only 3.6% of the variance in
sense of belonging was attributed to background and density variables. Appendix K displays the
regression statistics at each step. For the subsequent research questions, these demographic and
density variables were entered in step 1 and step 2 respectively, as a control in evaluating the
contribution of additional predictors beyond what is already afforded by these variables.
Research Question Two
RQ 2: To what extent do environmental, social, and behavioral factors contribute to the
prediction of off-campus students’ sense of belonging to the local community?
The second research question sought to determine if environmental, social, and
behavioral factors contributed to the prediction of sense of belonging. In the analysis, sense of
belonging (SoBT) was the dependent variable and the environmental factors construct (Factors)
two social factor constructs, neighborhood perceptions (Perceptions) and neighborhood relations
(Relations), and two behavioral constructs, positive behaviors (PosBehav) and negative
behaviors (NegBehav) were the predictor variables. The demographic and density variables used
as predictors in the first research question were entered in step 1 and step 2 respectively as
control variables in the model. The use of control variables allowed for understanding of how
the prediction changed with the addition of new variables, beyond the prediction associated with
the control variables. After the control variables were entered, the additional factors were
entered in the following order: step 3 as the environmental factors construct (Factors); step 4 as
the neighborhood perceptions construct (Perceptions), step 5 as the neighborhood relations
construct (Relations), and step 6 as both behavioral constructs (PosBehav, NegBehav). The
order of entry of the five predictor variables in steps 3 - 6 was chosen to reflect the way the

76

concepts are discussed in relation to one another in the sense of community work of Chavez and
Wandersman (1990). Additionally, the order of entry aligned with a similar approach taken by
McLaughlin (2011) during his dissertation research focused on off-campus students in the towngown context. Appendix L shows the results of the hierarchical regression analysis depicting the
contribution of environmental, social, and behavioral variables to predicting sense of belonging.
After demographic and density variables were entered as control variables in steps 1 and
2 respectively, 3.6% of the variance in sense of belonging was explained, although neither model
was statistically significant. After entry of Factors at step 3, the total variance explained by the
model was unchanged from the prior steps, as indicated by a R2 of .36 and a change in R2 of .000.
Model 3 was not statistically significant further signifying the Factors construct did not
reliability predict sense of belonging in the analysis.
After entry of Perceptions at Step 4, the model improved to predict 19.5% of the variance
in sense of belonging and was statistically significant, R2 = .195, F (15, 573) = 9.265, p < .001.
The Perceptions construct uniquely predicted 15.9% of the variance in sense of belonging, after
controlling for demographics, density, and Factors, R2 change = .159, F change (1, 573) =
113.198, p < .001. In Step 4, sex was the only significant individual predictor of sense of
belonging among the demographic characteristics. Sex (reference = female) was a negative
predictor of sense of belonging (b = -.090, p < .05) indicating males had less of a sense of
belonging to the local community than females. The Perceptions construct was a significant
positive predictor of sense of belonging (b = .413, p < .01) indicating those with more positive
perceptions of the local community had a higher sense of belonging.
After entering the Relations construct into the model at Step 5, the variance explained by
the model had negligibly changed to 19.6% from the earlier Step 4, yet the model remained

77

statistically significant overall, R2 = .196, F (16, 572) = 8.691, p < .001. The Relations construct
did not uniquely contribute to the variance in the Step 5 model, after accounting for the
predictive contributions of demographic, density, Factor, and Perceptions variables, R2 change =
.000, F change (1, 572) = .260, p = .610. The non-significant R2 change value at Step 5 signifies
the Relations construct did not reliability predict sense of belonging in the analysis. In Step 5,
significant predictors from prior steps were again significant including the Perceptions construct
(b = .432, p < .01) as a strong positive of belonging and Sex (male = 1) (b = -.089, p < .05) as a
modest negative predictor of belonging.
After the positive (PosBehav) and negative (NegBehav) behavior constructs were
entered in at Step 6, the variance explained by the model increased slightly to 20.3% and was
statistically significant overall, R2 = .203, F (18, 570) = 8.054, p < .001. The R2 change value was
not statistically significant, indicating the behavior constructs did not uniquely contribute to the
prediction of sense of belonging at this interval, R2 change = .007, F change (2, 570) = 2.58, p =
.077. Sex (male) (b = -.086, p < .05) and Perceptions construct (b = .425, p < .01) were again
significant predictor variables in the model, as in the previous two steps. Introduced in Step 6,
PosBehav was a statistically significantly predictor variable (b = .102, p <.05), while NegBehav
was not. The directionality of the standardized coefficients suggested being male was associated
with lower sense of belonging, while having positive perceptions of community and increased
participation in positive behaviors was associated with higher sense of belonging. The findings
associated with the PosBehav and NegBehav constructs should be interpreted with caution
because the unique contribution of the two behavior constructs at Step 6 were non-significant in
the overall, final model. In the final model, the most important predictor of sense of belonging
was the Perceptions construct accounting for 15.9% of the model variance as seen in Step 4.

78

Overall, the final model of environmental, social, and behavioral factors accounted for 20.3% of
the variance in sense of belonging.
To aid in interpretation of the findings of the hierarchical regression analysis of sense of
belonging in research question two, statistically significant predictors are summarized in Table 8.
Table 8
Summary of Statistically Significant Predictors of Sense of Belonging
Construct
Contribution to Sense of Belonging Step # Factor was Significant
Demographic variables

Sex (male)

4, 5, 6

Social factors (i.e.,
Perceptions, Relations)

Perceptions

4, 5, 6

Behaviors

PosBehav*

6

Note. * PosBehav finding should be interpreted with caution because R2 change was nonsignificant at the entry of positive and negative behavior constructs in Step 6.
Research Question Three
RQ 3: What is the relationship between sense of belonging in off-campus students and their
involvement in positive behaviors (e.g., neighborhood involvement, responsible neighbor) and
negative behaviors (e.g., alcohol violations, litter, fights) within the local community?
The third research question sought to understand the relationship between off-campus
students’ sense of belonging to the local community and their involvement in positive and
negative behaviors in the town-gown context. In contrast to the two prior research questions,
sense of belonging (SoBT) was a predictor variable in the regression equation, while positive
behaviors (PosBehav) and negative behaviors (NegBehav) were both criterion variables. With
two criterion variables and an interest in assessing the influence of sense of belonging on both
positive and negative behaviors, two separate hierarchical regression analyses were conducted.
In each model, demographic variables (Age, Sex, Race2, Race3, YrSchool, LiveOn, LiveOff,
RoomLog) and residential density variables (AptMed, AptLow, ResHigh, ResMed, ResLow)

79

were again entered in Step 1 and 2 respectively as control variables. The Factors, Perceptions,
and Relations constructs were entered in Steps 3 – 5, respectively. Sense of belonging (SoBT)
was entered in Step 6.
Hierarchical Regression Analysis of Positive Behaviors
Results of the hierarchical regression analysis depicting the contribution of demographic,
density, factors, perceptions, relations and sense of belonging constructs in predicting
participation in positive behaviors in the local community are depicted in Appendix M. After
Step 1, demographic characteristics contributed 8.5% of the variance in positive behaviors and
was statistically significant, R2 = .085, F (8, 580) = 6.769, p < .001. Significant, positive
predictors were RoomLog (b = .228, p < .001) and LiveOff (b = .147, p < .01). The positive
directionality indicated having higher numbers of roommates and living off campus more
semesters contributed to greater participation in positive behaviors.
Residential density predictors were entered in Step 2, which resulted in a statistically
significant increase in variance to 17.7% overall, R2 = .177, F (13, 575) = 9.531, p < .001. The
unique, statistically significant contribution of density predictors at this interval was 9.2%, R2
change =.092, F change (5, 575) =12.843, p < .001, which represented the largest interval
contribution in the model. In order of significance, positive predictors included ResHigh (b =
.226, p < .001), ResMed (b = .219, p < .001), ResLow (b = .141, p < .001), LiveOff (b = .116, p
< .05), and RoomLog (b = .093, p < .05). The Res variables represented residential areas of
houses characterized as high, medium, and low levels of density. With all three levels of
residential density being significant, the finding suggested living in a house, regardless of
density, was associated with increased participation in positive behaviors. As density levels
increased so did the contribution of the variable to positive behaviors participation.

80

The third step introduced the environmental factors construct (Factors) into the model,
which resulted in a statistically significant increase in variance to 18.6% overall, R2 = .186, F
(14, 574) = 9.394, p < .001. The unique contribution of the Factors construct at this interval
minimal, but statistically significant, at less than 1%, R2 change =.009, F change (1, 574) =
6.447, p < .01. In order of significance, positive predictors were ResMed (b = .223, p < .001),
ResHigh (b = .214, p < .001), ResLow (b =.146, p < .001), Factors (b =.098, p < .01), AptMed
(b = .047, p < .05), and LiveOff (b =.037, p < .05). All three Res density variables and LiveOff
were again indicative of increased participation in positive behaviors. Present in this step only,
AptMed was a significant indicator suggesting medium density apartment dwellers were more
involved in positive behaviors than other apartment dwellers. The positive directionality of
Factors suggested as perceptions of environmental problems increased, so does participation in
positive behaviors.
After entering the Perceptions construct into the model at Step 4, the variance explained
by the model increased to 24.9%, which was statistically significant, R2 = .249, F (15, 573) =
12.645, p < .001. The statistically significant Step 4 interval contributed 6.2% to the model, R2
change =.062, F change (1, 573) = 47.497, p < .001. In order of significance, familiar positive
predictors were ResMed (b = .208, p < .001), ResHigh (b = .189, p < .001), ResLow (b = .152, p
< .001), Factors (b = .106, p < .01), and RoomLog (b = .091, p < .05). The newest addition
Perceptions (b = .259, p < .001) was the largest significant predictor at Step 4. The positive
directionality of Perceptions suggested as positive perceptions of the local community increased,
so did participation in positive behaviors.
The fifth step introduced the Relations construct to the model, which improved the model
predictability to 28.7% of the variance in positive behavior participation, R2 = .287, F (16, 572) =

81

14.406, p < .001. The unique, statistically significance of the fifth interval contributed 3.9% to
the model variance, R2 change =.039, F change (1, 572) = 30.923, p < .001. In order of
significance, familiar positive predictors were ResMed (b = .198, p < .001), ResHigh (b = .165, p
< .001), ResLow (b = .150, p < .001), and Factors (b = .093, p < .01). Relations (b = .284, p <
.001) was the largest significant predictor at this interval with the positive directionality
suggesting as neighborhood relations (e.g., social interactions, casual conversation) increased,
participation in positive behaviors also increased.
The sixth and final step was the addition of Sense of Belonging (SoBT) to the regression
model. The overall model was statistically significant and explained 29.4% of the variance in
positive behavior participation, R2 = .294, F (17, 571) = 13.958, p < .001. The unique
contribution of SoBT at the sixth interval was statistically significant yet minimal, indicating
SoBT contributed less than 1% to the variance in positive behavior participation, R2 change =
.006, F change (1, 571) = 5.123, p < .05. Significant predictors familiar from prior steps
included Relations (b = .287, p < .001), ResMed (b = .197, p < .001), ResHigh (b = .164, p <
.001), ResLow (b = .148, p < .001), and Factors (b = .091, p < .01). Additionally, SoBT (b =
.089, p < .05) was a modest positive predictor, suggesting higher feelings of belonging were
associated with increased participation in positive behaviors. In the final model, Relations (b =
.287, p < .001) was the strongest predictor, while addition of the density variables at Step 2
showed the largest incremental increase of the model at 9.2%. Overall, the final model
accounted for 29.4% of the variance in positive behavior participation.
Hierarchical Regression Analysis of Negative Behaviors
The results of the hierarchical regression analysis depicting the contribution of
demographic, density, Factors, Perceptions, Relations and Sense of Belonging constructs to

82

predicting participation in negative behaviors (NegBehav) are presented in Appendix N. At the
end of Step 1, demographic characteristics had contributed 14.6% of the variance in negative
behavior participation and was statistically significant, R2 = .146, F (8, 580) = 12.423, p < .001.
Three predictor variables were statistically significant predictors with positive directionality.
RoomLog (b = .271, p < .001) was the largest predictor suggesting a higher number of
roommates was associated with greater participation in negative behaviors. Sex (male = 1) (b =
.191, p < .001) suggested males participate in negative behaviors more than females. LiveOff (b
=.137, p < .01) showed the more semesters a student lived off campus, the more their
participation in negative behaviors increased.
After residential density variables were entered in Step 2, the overall model was
statistically significant and accounted for 19.5% of the variance in negative behavior
participation, R2 = .195, F (13, 575) = 10.742, p < .001. The unique contribution of the density
variables entered at Step 2 was 4.9% of the variance in the total model and was statistically
significant, R2 change = .049, F change (5, 575) = 7.022, p < .001. Four predictor variables were
statistically significant with positive directionality. RoomLog (b = .206, p < .001) was again the
largest predictor as described in the previous step. ResHigh (b =.198, p < .001) was the next
largest contributor suggesting living in a high-density area of residential homes was associated
with greater participation in negative behaviors, while levels of lower density homes were not
predictors of negative behavior participation. Sex (male =1) (b = .194, p < .001) and LiveOff (b
=.124, p < .05) were again positive predictor variables.
After entering the Factors construct into the model at Step 3, the variance explained by
the model increased to 21.7%, which was statistically significant, R2 = .217, F (14, 574) =
11.382, p < .001. The unique statistically significant contribution of the Factors interval was

83

2.2% of the variance, R2 change = .022, F change (1, 574) = 16.042, p < .001. In order of
significance, positive predictors familiar from prior steps were Sex (male = 1) (b = .201, p <
.001), RoomLog (b = .192, p < .001), ResHigh (b = .179, p < .001), LiveOff (b = .107, p < .05),
and LiveOn (b = .080, p < .05). The Factors construct (b = .152, p < .001) was a statistically
significant positive predictor indicating as perceptions of environmental problems increased,
participation in negative behaviors increased.
After the Perceptions construct was entered at Step 4, the variance explained by the
model increased to 28.4% and was statistically significant, R2 = .284, F (15, 573) = 15.188, p <
.001. The unique contribution of the interval was statistically significant adding 6.7% to the
model variance, R2 change = .067, F change (1, 573) = 53.810, p < .001. The Perceptions
construct (b = .269, p < .001) was a significant positive predictor indicating as positive
perceptions of the community increased, participation in negative behaviors also increased.
Significant positive predictors from prior steps included RoomLog (b = .199, p < .001), Sex
(male = 1) (b = .186, p < .001), Factors (b = .160, p < .001), and ResHigh (b = .153, p < .001).
At Step 5, the Relations construct was added to the model, which increased the variance
explained to 37% and was statistically significant, R2 = .370, F (16, 572) = 20.996, p < .001. The
unique statistically significant contribution of the Relations interval was 8.6%, R2 change = .086,
F change (1, 572) = 77.646, p < .001. In order of significance, positive predictors familiar from
prior steps were Sex (male = 1) (b = .170, p < .001), RoomLog (b = .169, p < .001), Factors (b =
.141, p < .001), and ResHigh (b = .118, p < .001). Relations (b = .424, p < .001) was the largest
significant positive predictor at Step 5, suggesting as social interactions with neighbors increased
so did participation in negative behaviors.
With the introduction of Sense of Belonging (SoBT) in the sixth and final step, the

84

variance explained by the model remained unchanged at 37%, but was statistically significant, R2
= .370, F (17, 571) = 19.728, p < .001. The unique contribution of SoBT at Step 6 was not a
statistically significant contribution to the model variance. In the final model, significant
positive predictors familiar from prior steps included Relations (b = .424, p < .001), Sex (male =
1) (b = .171, p < .001), RoomLog (b = .169, p < .001), Factors (b = .141, p < .001), and ResHigh
(b = .118, p < .001). Additionally, SoBT (b = .006, p = .878) was a non-significant predictor,
which suggested sense of belonging did not reliably influence participation in negative activities.
In the final model, Relations (b = .424, p < .001) was the strongest predictor of participation in
negative behaviors. The final model of demographics, density, Factors, Perceptions, Relations
and Sense of Belonging accounted for 37% of the variance in negative behavior participation.
Summary of Significant Predictors of Student Behaviors
To compliment the findings presented for the hierarchical regression analyses of positive
and negative behaviors in research question three, Table 9 summarizes statistically significant
predictors for both behavior types.
Table 9
Summary of Statistically Significant Predictors of Positive and Negative Behaviors
Factor
Positive Behaviors
Negative Behaviors
Demographic variables
• Semesters lived off campus
• Semesters lived off campus
• Number of roommates
• Number of roommates
• Sex (male)
Density categories
• Predictive power of houses at • High density houses
all density levels; no
apartment levels
Environmental factor
• Perceptions of environmental • Perceptions of environmental
problems
problems
Social factors (i.e.,
• Relations, not Perceptions
• Relations, not Perceptions
Perceptions, Relations)
Sense of Belonging
• Small, positive contribution
• No contribution of Sense of
of Sense of Belonging
Belonging

85

Summary
The current chapter began with details of data screening efforts, sample demographic
composition, and construct development. Hierarchical regression analysis was reviewed and
regression assumptions were tested and met. Regression analysis conducted for the research
questions revealed demographic variables alone did not significantly contribute to the prediction
of sense of belonging in off-campus students. The addition of environmental, social, and
behavior factors in a sequential model showed the neighborhood perceptions factor and positive
behavioral participation factor contributed significantly to the prediction of sense of belonging
more than other factors considered. Regression findings also revealed sense of belonging was a
statistically significant predictor of positive behavior participation only, albeit a modest one,
suggesting other factors must also influence both positive and negative behavior participation in
the town-gown context. Chapter 5 shares the discussion of these data and how they relate to
literature on sense of belonging, commuter students, and the town-gown context. Implications
for research and practice are presented, as well as future directions for research.

86

CHAPTER FIVE
Discussion
The current study examined the sense of belonging off-campus students felt toward their
local community of residence during college. Central to the study was exploration of the
environmental, social, and behavioral factors that contribute to sense of belonging, and in turn
how these factors and feelings of belonging served to predict participation in positive and
negative behavior within the town-gown context. This chapter summarizes the results and
explains the findings from hierarchical regression analyses in association with literature on
commuter students, sense of belonging, and town-gown relations. The research questions built
sequentially on one another, so the discussion reflects this sequence in presentation. The
contributions and implications of the current study to practice and policy are discussed. The
chapter concludes with directions for future research.
Motivation for the Current Study
The impetus for the current study emerged from the need to address behavior issues in
the town-gown community and the practical need to know more about off-campus students as
central actors in these behaviors. The general lack of knowledge about off-campus students was
due to two primary factors. First, the broad, historical definition of commuters has included all
students who do not live in university sponsored housing, regardless of within-group differences
that likely exist (Dugan et al., 2008). As such, off-campus students are considered commuters
and limited attention is paid to them as a unique subpopulation in higher education. Second,
contemporary town-gown literature and news media focused on student behaviors as a prevalent
challenge in the town-gown community (Baker-Minkel et al., 2004; Twohey, 2007), yet failed to
formally examine off-campus students as central actors in these behaviors. Existing town-gown

87

research has focused largely on university administrators and municipal leaders (Bruning et al.,
2006), and non-student residents (Kittle, 2005) as stakeholders in the town-gown community,
with little consideration of off-campus students in the same community context (McLaughlin,
2011). Thus, the current study focused on off-campus students explicitly within the local
community context.
Equally important, there was a need to address behavior concerns central to town-gown
tensions. The theory of sense of belonging provided a way to conceptualize the current study, as
the need to belong is known as a pervasive driving source of motivation capable of eliciting
outcomes including engagement, achievement, retention, persistence, optional functioning, and
well-being (Hausmann et al., 2009; Rhee, 2008; Strayhorn, 2012). As a motivator of outcomes,
could sense of belonging be a source of influence on behavioral choices in off-campus students?
The current study explored this consideration, and in turn, evaluated environmental, social, and
behavior factors that influenced sense of belonging.
Sample Profile
Attention to the demographic characteristics of the sample in the current study was
warranted because off-campus students have been largely absent from town-gown research and
subsumed in the vast, homogenous definition of commuters. Reliance on an antiquated
definition of commuters mistakenly personifies commuters as either “townies” residing with
their parents in the local community, or as non-traditional aged students who often live a
considerable distance from campus and must balance education with competing life roles such as
employment and family commitments (Jacoby, 2000a; Stewart & Rue, 1983). Sample
demographics indicated off-campus students in the study differed from traditional commuter
characteristics in several key ways.

88

Study respondents were traditional-aged students instead of non-traditional aged as
commuter literature often depicts (Jacoby, 1989). Most students had lived on-campus before
moving off-campus to the local community, which challenges the townie personification of
living at home with one’s parents and the notion commuters often live a considerable distance
from campus, instead of locally (Jacoby, 2000a). Most participants lived with other students as
roommates, which further challenges the townie personification and the notion commuters are
balancing family responsibilities in their place of residence (Jacoby, 1989; Stewart & Rue,
1983). These demographics depict where off-campus students differ from the traditional
conceptualization of commuters, while also revealing the similarity off-campus students share
with their residential peers, who are also traditional-aged students with roommates (Hintz, 2011).
Dugan et al. (2008) argued for this need to explore subpopulations of commuter students and
residential students in greater depth, rather than continuing to rely on the residential experience
in higher education as a normative benchmark for all students.
As the sample demographics illuminated departures from the traditional commuter
student definition, the same demographic details affirm the descriptive terms used to characterize
off-campus students broadly within town-gown literature and popular news media. Off-campus
students are described as traditional-aged undergraduates (i.e., 18 - 22 years old) who have lived
on-campus in residence halls before moving off-campus to the local community (Hintz, 2011).
Most traditional off-campus students rent, live among dense enclaves of fellow students, and live
independently of their guardians and direct supervision (Gumprecht, 2003; McLaughlin, 2011).
The sample in the current study reflected this contemporary perception of off-campus students,
thus providing empirical confirmation of what is known subjectively in town-gown practice.

89

The demographic composition of the current sample confirmed the characterization of
off-campus students in the town-gown context. At the same time, the composition demonstrated
the need for consideration of within-group differences in the commuter population, including
identification of unique subpopulations within the broader commuter label. The current study
confirmed and strengthened initial understandings of off-campus students in the town-gown
context, while adding off-campus students as a population for further consideration in higher
education research and practice.
Sense of Belonging in Off-Campus Students
The current study demonstrated background characteristics (age, sex, race, year in school,
semesters lived on and off campus, number of roommates) of off-campus students were not
significant predictors of sense of belonging and overall, demographics account for only 2% of
the variance in belonging. These findings mirror what McLaughlin (2011) found when assessing
the influence of demographic characteristics on sense of belonging for student residents of a
near-campus neighborhood at Ohio State University. Similarly, Johnson et al. (2007) found
demographic variables did not predict sense of belonging in their study of students of Color, with
the only exception being gender as a predictor of belonging for Latino students.
In addition to background characteristics, the influence of residential density levels on
sense of belonging was explored. Somewhat surprisingly, density levels were not significant
predictors of sense of belonging; accounting for only 1% of the variance in sense of belonging.
McLaughlin (2011) reported having neighborhood relationships or social ties increased feelings
of belonging, so one might have reasonably anticipated living in an area of greater density of
people would provide the opportunity to potentially develop a greater number of relationships,
thus cultivating a greater sense of belonging. The current density finding does not support this

90

line of thinking but instead calls to question potential differences in the influence of density
based on relationships or social ties and density based on physical quantity of people within an
area. The predictive capacity of social ties on sense of belonging suggests density of
relationships may be more valuable in cultivating belonging than density based on quantity of
individuals in a specified unit of space. Further exploration of the role of density is warranted to
evaluate the density finding of the current study.
Factors Influencing Sense of Belonging
The work of town-gown administrators often focuses on increasing the sense of
belonging of off-campus students as a means of cultivating greater civility and better behavior in
the local community context (Whitson, 2011/2013). Identifying factors to further enhance
strategies used to promote sense of belonging was the stimulus behind evaluating the
effectiveness of environmental, social, and behavior factors in influencing belonging. The
factors explored (i.e., Factors, Perceptions, Relations, PosBehav, NegBehav) mirrored the factor
components theorized by McMillan and Chavis (1986) and confirmed in the model of sense of
community belonging by Chavis and Wandersman (1990). The items for Perceptions and
Relations constructs were borrowed from McLaughlin (2011) because they aligned with the
intention of the model of sense of community belonging (Chavis & Wandersman, 1990), yet
offered language tailored to a town-gown context specifically.
Social Factors
The most important positive predictor of sense of belonging was the neighborhood
perceptions factor (Perceptions), which demonstrated that perceptions or thoughts about one’s
neighborhood and neighbors, not necessarily the reality of the context or relationships,
influenced sense of belonging. This finding corroborated the work of Johnson et al. (2007),

91

which concluded students’ perceptions of the residence hall environment as socially supportive
was a positive predictor of sense of belonging. The findings substantiated similar findings about
neighborhood perceptions in work of Chavis and Wandersman (1990) and McLaughlin (2011).
The neighborhood relations factor (Relations) was not a significant predictor of
belonging, which indicated actions (e.g., borrowing a tool) and social interactions (e.g., casual
social visits) within the community did not influence belonging. In contrast, McLaughlin (2011)
found the same Relations construct was the most significant, positive predictor of sense of
belonging in his model and Chavis and Wandersman (1990) also documented a positive,
significant finding for neighborhood relations. As such, the findings of the current study should
be viewed with caution since they do not align with existing research. One possible explanation
of the discrepancy lies in the difference of the community context and associated participants
used in each study (McMillan & Chavis, 1986). The works of McLaughlin (2011) and Chavis
and Wandersman (1990) focused on analysis of residents of a designated neighborhood, where
the current study included student residents of the City of East Lansing broadly and referenced
their neighborhood of residence but not a consistent neighborhood. This difference suggested
sense of belonging to community may be location or context specific and perhaps better used in
smaller scale contexts instead of community level evaluations.
Environmental Factor
The theory of sense of community articulated the importance of perceptions of problems
in the environment (e.g., house conditions, litter, crime) in influencing feelings of community
belonging. In the corresponding model developed and tested for sense of community belonging,
the environmental factor was a non-significant predictor of sense of community belonging
(Chavis & Wandersman, 1990). The current study confirmed the same finding as the

92

environmental factor was a non-significant predictor and did not contribute to the overall model
of sense of belonging in off-campus students.
Demographic and Density Characteristics
Density and demographic characteristics, except for sex, were non-significant predictors
in the model, which substantiated findings of McLaughlin (2011) about the role of background
variables in his belonging model. For sex, the significant finding indicated males had a lower
sense of belonging than females, which contrasted McLaughlin’s (2011) finding that showed
female student had less of a sense of belonging than males. One possible explanation for the
discrepancy was McLaughlin’s (2011) model of sense of belonging included a regression block
about perceptions of crime and safety within the neighborhood, whereas the current study did not
include this focus. McLaughlin (2011) attributed the directionality of his finding to perceptions
that females had a lower sense of belonging because they felt less safe in the neighborhood. The
current finding does not support or refute this claim but does showcase how additional variables
in a model can influence the significance and directionality of one another.
Positive and Negative Behavior Factors
The final factors considered for their influence in predicting sense of belonging were the
positive and negative behavior constructs. Created intentionally for the current study, the
constructs aimed to ascertain the influence of participation in positive and negative behaviors in
predicting sense of belonging. Participation in positive behaviors (e.g., attending neighborhood
events, use of community services) was a significant positive predictor of belonging, which
showed greater levels of participation in positive behaviors elicited greater feelings of belonging
to the local community. Similarly, McLaughlin (2011) found participation in community
building activities in a near-campus neighborhood was a significant, positive predictor of sense

93

of belonging in off-campus students. In multiple studies Strayhorn (2012) showed first-year
students, Black male students, and students of Color increased their sense of belonging through
involvement in positive activities, like student clubs and educationally purposeful activities, such
as peer mentoring and living-learning communities. In contrast, participation in negative
behaviors was not a significant predictor of sense of belonging. No additional empirical work
was found to support or negate this finding, so it should be cautiously interpreted as an initial
contribution regarding negative behavior participation and sense of belonging.
Exploration of Student Behaviors in a Town-Gown Context
The former half of the current study focused on gauging the sense of belonging of offcampus students and exploring factors that influenced this belonging. The latter half focused on
investigating the role of demographics, density, environmental factors, social factors, and sense
of belonging in predicting participation in positive and negative behaviors. Chavis and
Wandersman (1990) advanced the premise that variables of influence in the model of sense of
community belonging were dynamic and bidirectional in nature, demonstrating sense of
belonging was “both a cause and effect in local action” in the community context (p. 73).
Additionally, in the theory of belonging, Baumeister and Leary (1995) posited the innate human
desire for belonging was sufficient to motive behavior. Subsequent research demonstrated the
influence of sense of belonging in higher education on outcomes including engagement,
achievement, retention, persistence, optional functioning, and well-being (Hausmann et al., 2009;
Rhee, 2008; Strayhorn, 2012). The current study extended this exploration of sense of belonging
as a motivator into the town-gown context by examining positive and negative behaviors as
discussed below. The nature of this inquiry was exploratory more than confirmatory, as little is
known about student behaviors in the town-gown context beyond what is portrayed through

94

limited literature available and the contextual and subjective relevance of town-gown practice
(Hintz, 2011; Twohey, 2007).
Influences on Positive Behavior
Town-gown administrators expend effort to encourage positive community behavior with
hope of cultivating greater sense of civic responsibility and community belonging in off-campus
students. In turn, they hope feelings of greater belonging and care in the community translate
into less involvement in negative or problematic behaviors in the local community. To this end,
the current study investigated the influence of various factors in predicting positive behavior
participation in off-campus students.
Density levels associated with living in a house emerged as a consistent positive
predictor of participation in positive behaviors across all steps of the regression model, while
density levels associated with apartment living were not significant at any step. This contrast
raised a question about the nature and influence of density based on unit type (i.e., houses and
apartments). Density in the current study was a calculation based on the number of individuals
per unit of space, yet the finding suggested density, regardless of level (i.e., high, medium, low)
did not equate to the same outcome in both unit types. Could density be experienced or
perceived differently by the dwellers of each unit type? McLaughlin (2011) found simply having
friends in one’s neighborhood was not a predictor of neighborhood outcomes, where as having
friendships that resulted from neighboring (networks of place) was a positive predictor of
participation in a community building program and a positive predictor of social ties in the
community. To parallel this thinking, suppose houses at all density levels allowed students to
have greater numbers of interactions with their neighbors compared to the experience of
apartment dwellers at comparable density levels, thus resulting in the predictive power of all

95

three house density variables. Another explanation for the different predictive influence of
density in houses and apartments is opportunities for participation in positive behaviors were not
equally available to dwellers of both unit types. The current study did not evaluate the frequency
of offerings or the availability of opportunities, so the possibility of inequity is offered as an
exploratory explanation, not a conclusive one.
Regardless of the unit type, the current study illustrated the role neighbors and neighborly
relations had in influencing participation in positive behaviors in the local community. In early
stages of the regression model, the neighborhood perceptions construct showed off-campus
students’ positive perceptions of their neighbors’ friendliness, trustworthiness, and reliability,
were associated with higher participation in positive behavior. Similarly, the neighborhood
relations construct demonstrated increasing interaction between neighbors (e.g., loaning or
borrowing a tool, helping with favors, casual social interactions) yielded higher participation in
positive behaviors in the local community. Interestingly when the neighborhood relations
construct was added to the model, neighborhood perceptions became non-significant, which
suggested actual social interactions between neighbors were more predictive than conceptual
perceptions of neighborly relations in cultivating positive behavior participation in off-campus
students. The current finding substantiated Chavis and Wandersman’s (1990) finding that
neighboring relations, a construct expressed as friendships and levels of social engagement
among neighbors, was a strong positive predictor of participation in a neighborhood association
as the behavior they sought to elicit in neighbors.
Further evidence of the importance of peers and others in encouraging participation in
positive behaviors was seen in the significant positive predictive power of number of roommates
and number of semesters lived off-campus. In both instances, greater numbers of roommates and

96

more semesters lived off campus predicted greater frequency of participation in positive
behaviors. This finding suggested having more roommates and living off-campus longer might
afford off-campus students more opportunity to develop friendships and peer relationships,
which in turn, exposed students to greater opportunities for participation in positive behaviors.
The power of peer relationships in exposing students to diverse opportunities is documented in
the work of Astin (1977/1993) and Pascarella and Terenzini (2005), as well as findings from the
National Survey of Student Engagement (2013).
Moving beyond neighborly interactions, perceptions of environmental problems was a
positive predictor of positive behavior participation. This finding illustrated off-campus students
who perceived a greater sense of problems (i.e., traffic, litter, crime) were more likely to engage
in positive behaviors in the local community. As off-campus students’ perceptions of
environmental concerns grew, were they spurred to participate in positive behaviors (e.g., use
community resources and services, attend community events) in the local community?
Seemingly contradictory, this notion aligned with Chavis and Wandersman’s (1990) finding
showing participants utilized community resources and were empowered to join a community
organization as strategies to alleviate perceived neighborhood problems.
Lastly, sense of belonging emerged as a small, yet significant positive predictor of
participation in positive behaviors in the local community. The smallness of the contribution
demonstrated sense of belonging was not a crucial source of motivation alone for participation in
positive behaviors, but rather interacted with other factors to influence behavior. Together,
demographic characteristics, density levels, environmental factors, social factors, and sense of
belonging predicted 29% of the variance in positive behavior participation in off-campus
students. Strayhorn (2012) reached the same conclusion when discussing the success of graduate

97

students in reaching academic goals (i.e., dissertation completion, research projects) as a blend of
factors including sense of belonging, socialization efforts, and student-faculty interactions.
Influences on Negative Behavior
Much the same with efforts to cultivate positive behavior participation, designing
strategies to reduce negative behaviors in the local community were central to the work of towngown administrators. The majority known about student behavior concerns in the town-gown
context comes from news media, public opinion, and practical and professional knowledge of
town-gown administrators. Thus, gaining insight into factors motivating off-campus students’
participation in negative behaviors provides empirical data to guide refinement of town-gown
efforts to reduce the frequency and severity of negative behaviors in the local community.
As found with positive behaviors, density was again a significant positive predictor of
participation in negative behaviors in the local community. The difference being only high
residential density (i.e., houses, not apartments) emerged as a positive predictor, no other levels
of residential or apartment density were significant. This finding furthered the earlier notion
dense housing areas were conduits for friendship development and subsequent exposure to
negative behavior opportunities. Within the East Lansing context, large parties, excessive noise,
and raucous student behaviors often originate in student-dominated neighborhoods near campus.
A disproportionate number of calls city-wide for police, parking enforcement, and emergency
services are drawn to these areas in response to negative behaviors (S. Webster, personal
communication, April 5, 2017). In corroboration, areas of dense enclaves of students living sideby-side were noted as prominent hotbeds of town-gown behavioral concerns by Gumprecht
(2008), Hintz (2011), and McLaughlin (2011). Additionally, greater perception of these
problems (e.g., noise, parties, parking enforcement) in the environment was predictive of

98

negative behavior participation. This predictive relationship suggests greater awareness of these
problems was sufficient to normalize participation in the same negative behaviors perceived in
the community.
Several demographic characteristics were significant positive predictors of participation
in negative behaviors in the current study, including sex, age, number of roommates, and
semesters lived on-campus and off-campus. Males were more likely than females to participate
in negative behaviors, which supported research from Kaplowitz and Campo (2004) that showed
males were more likely than females to participate in riots, consume alcohol, and condone highrisk behaviors in a college environment. Having higher numbers of roommates and living offcampus longer were both associated with greater likelihood of negative behavior participation,
perhaps another indication of the power of peer relationships to expose students to opportunities
in the community, albeit negative behaviors instead of positive ones. Similarly, older students
were more likely than their younger peers to engage in negative behaviors, which may reflect
greater peer exposure to negative behaviors or more time to develop familiarity with
opportunities for participation in negative behaviors.
Interestingly, number of semesters lived on-campus was also a positive indicator of
negative behavior participation in the local community. One potential explanation for this
finding lies in the transition from the structure of on-campus living to the freedom of off-campus
living. Frequently, MSU undergraduates move off-campus after fulfilling the first-year campus
residency requirement in pursuit of independent living and freedom from perceived restrictions
of residence hall rules (S. Webster, personal communication, April 5, 2017). Hintz (2011)
argued this independent living results in town-gown behavior concerns because most off-campus

99

students “possess limited knowledge of the rights and responsibilities of community living with
little oversight” (p. 87).
Neighborhood perceptions and neighborhood relations again emerged as positive
predictors of negative behaviors, as found earlier with positive behavior participation. This
finding furthered the earlier argument that neighborly influence, either in interactions or
perceptions of belonging, was sufficient to drive participation in behaviors, both positive and
negative. In the current study, the power of peer influence was stronger in predicting negative
behaviors (B = .424) than positive behaviors (B =.287) in the local community. As McLaughlin
(2011) explained, perhaps finding a party to attend, alcohol to consume, and mischievous
behavior to engage in was easier with friends and neighbors in a university-community context.
Lastly, sense of belonging proved not to be a significant step in the model, nor a
significant predictor of negative behavior participation. In one way, this finding conflicted with
a fundamental town-gown strategy aimed at cultivating community belonging to induce greater
caring and civility thereby reducing negative behavior participation. Equally important, sense of
belonging was not a reliable predictor of negative behavior participation, suggesting town-gown
administrators might not have to worry efforts to increase sense of belonging to the local
community might also elicit increases in negative behavior participation. Instead, this finding
illustrated other variables intervened with greater influence in promoting negative behaviors
participation in the town-gown context. Likewise, Chavis and Wandersman (1990) depicted
sense of community belonging as being interactional and bidirectional in nature, suggesting
interplay between variables was dynamic much like the many influences present in community
living. This bidirectional premise was upheld by Strayhorn (2012) who described a “chicken
before the egg scenario” in reference to his suggestion sense of belonging can influence an

100

outcome of interest, but the reverse may also be true (p. 102). Considering the exploratory
nature of the current study, the non-significance of sense of belonging in predicting negative
behaviors should be considered with caution and as an indication further exploration is needed.
Comparing Positive and Negative Behaviors
Having examined predictors of each behavior type in the two prior sections, attention is
now turned to considering influences on both behavior types in tandem. Analyzing positive and
negative behaviors alongside one another provided an opportunity to compare and contrast the
predictors associated with each behavior type. This broader view of both behavior types helps to
illustrate patterns of association among predictors and behaviors.
In considering the influence of predictor variables, a Venn diagram is a useful concept to
visualize the overlap and divergence of predictor variables in relation to one another and their
association with behavior types. Figure 2 is a Venn diagram representing the associations of the
neighborhood perceptions, neighborhood relations, and sense of belonging predictors in research
question three. The areas of overlap in Figure 2 are labeled to correspond to the behavior type
predicted by the indicated predictor variables in the regression analysis of student behaviors.

Neighborhood
Perceptions
B
A
Sense of
Belonging

Legend

Neighborhood
Relations

Positive Behaviors

A

Negative Behaviors B

Figure 2. Venn Diagram of Predictors Associated with Positive and Negative Behaviors

101

The interpretation of Figure 2 provides an important observation about the influence of
the predictor variables in predicting positive and negative behaviors. Segment A shows the joint
influence of neighborhood perceptions, neighborhood relations, and sense of belonging predicted
positive behaviors, while segment B shows the presence of neighborhood perceptions and
neighborhood relations, without sense of belonging, predicted negative behaviors. All things
being equal, the presence of sense of belonging in the regression analysis was associated with
positive behaviors, while the absence of sense of belonging was associated with negative
behaviors. This notion shows neighborly interactions (i.e., relations and perceptions) between
neighbors are associated with both behavior types, while the addition of belonging and care in
the interactions predicts positive behaviors.
Baumeister and Leary (1995) offer insight into the role of belonging in behaviors with
their discussion of human interactions without an ongoing bond of caring. Their work indicates
people have a need for relatedness and mutual caring in interactions sustained over time. The
authors argue interactions absent relatedness and caring leave individuals dissatisfied and
yearning for belonging. Baumeister and Leary (1995) demonstrate the presence of belonging in
interactions is associated with positive outcomes, such as happiness, health, and general welfare,
while the absence of belonging in interactions is associated with outcomes including depression
and loneliness.
To parallel this thinking, neighborly interactions characterized by feelings of belonging
and care predict positive behavior outcomes, while interactions alone predict negative behaviors.
For town-gown administrators, this notion justifies why efforts to cultivate belonging are
worthwhile and necessary if positive behaviors are the desired outcome of neighborhood
programming. Town-gown programs that stop short of cultivating belonging and focus instead

102

on promoting neighborly interactions alone are linked to negative behaviors instead of the
desired positive behaviors.
Contributions and Implications of the Current Study
The current study drew from three distinct areas of higher education scholarship, namely
literature on commuter students, town-gown relations, and sense of belonging in college.
Together, the literature dovetailed with practice to provide the structure and support for the
current study dedicated to the exploration of off-campus students’ sense of belonging to the local
community. Contributions and implications are offered in line with these three areas of literature
and practice.
Contributions to Theory and Research
In higher education, the term commuter student has long been known as broadly inclusive
of students based on a shared commonality of not living in university-owned housing (Jacoby,
1989; Stewart & Rue, 1983) while simultaneously being limiting due to a lack of recognition of
the diversity inherent within the commuter population (Dugan et al., 2008; Jacoby, 1989). In
response to this limitation, the current study advanced exploration of the within-group diversity
of the commuter population by focusing empirical attention on off-campus students explicitly.
Demographic characteristics of the sample validated off-campus students were distinct from the
antiquated, homogenous conceptualization of commuters and instead shared similarity with their
residence hall peers. These results demonstrated the need for nuanced attention to off-campus
students and other subpopulations subsumed within the commuter population. Research focused
on these subpopulations would further reveal within-group differences in the commuter
population and inform the design of higher education services and resources tailored to these
subpopulations. Additionally, similarities found between off-campus students and residence hall

103

students expose a need to delve deeper in comparative analysis of the two groups. Commuter
literature is replete with studies comparing the college experience and associated outcomes of
commuter students with those of residential students. In consideration of this existing literature,
growing acknowledgment of the diversity inherent in the commuter population suggests a need
to revisit past research to ascertain if conclusions reached initially remain true for the commuter
population and its many subpopulations.
Within higher education research, sense of belonging was used to study various student
cohorts including students of Color (Hausmann et al., 2009; Hurtado & Carter, 1997; Johnson et
al., 2007; Strayhorn, 2012), residential students (Johnson et al., 2007), first-year students
(Hoffman et al., 2002), graduate students (Strayhorn, 2012), and part-time students (Kember,
Lee, & Li, 2001). The current study contributed to this established body of student-focused
belonging literature by introducing off-campus students into sense of belonging research.
Adding a new student cohort expands the inclusivity of belonging research and provides further
opportunity to explore and refine understanding of the role sense of belonging plays in eliciting
student outcomes in a diverse student population. The local community context of the current
study also extended belonging research beyond campus boundaries, as prior research focused on
classroom and campus belonging (Freeman et al., 2007), residence hall belonging (Johnson et al.,
2007), and university belonging generally (Hurtado & Carter, 1997; Rhee, 2008). This extension
beyond campus boundaries acknowledges the reality that for off-campus students, the local
community is a significant part of their collegiate experience as their place of residence during
college. Further research should delve into the effects the off-campus environment has on
college outcomes of the students who live there.

104

Much of existing town-gown literature has focused on organizational level interactions to
navigate shared opportunities and challenges in the university and local municipality context
(Kemp, 2013). Prior research has examined areas including economic development (O’Mara,
2012; Porter & Grogan, 2002/2013), public service demands (Martin & Samels, 2006/2013;
Nichols, 1990), and amenities such as arts, culture, and entertainment (Baker-Minkel et al., 2004;
Warfield, 1995). The current study drew attention to the student experience in the town-gown
context and focused on student behaviors as a central component of town-gown tensions. Insight
into off-campus students’ participation in positive and negative behaviors provides initial data to
drive further inquiry and exploration aimed at reducing student behavior concerns that elicit
town-gown tensions. The current study also adds off-campus students to town-gown literature, as
most prior studies have focused on other stakeholder groups including university administrators
and municipal leaders (Bruning et al., 2006) and non-student residents (Kittle, 2005). The
current study examined off-campus students’ feelings of belonging, factors influencing this
belonging, and their participation in various behaviors in the local community. These findings
are offered as a starting point and underscore a need for further exploration of the off-campus
student experience.
Lastly, the current study employed a novel approach to studying student populations in
higher education, namely the use of Geographic Information System (GIS) mapping to ascertain
measures of student density in living environments. Using GIS mapping, density categories (i.e.,
high, medium, low) based on unit types (i.e., houses, apartments) were created and subsequent
analysis considered the influence of density in predicting sense of belonging and student
behaviors in off-campus students. Although a first effort, the potential to gain insight into the
role density plays in a myriad of student outcomes is immense.

105

Contributions to Practice and Policy
The conceptualization for the current study emerged from the researcher’s experience as
a student affairs professional engaged in town-gown relations. With a basis in practice, there are
several contributions and implications from the current study meant to inform practice. For
student affairs and town-gown professionals working with off-campus students, there has long
been a lack of empirical data about off-campus students (Kemp, 2013). The current study shed
light on off-campus students to both address this absence and to provide needed insight into the
student population central to town-gown relations work. Demographic characteristics collected
in the study provide an initial descriptive profile of off-campus students, which expands and
compliments anecdotal and subjective estimations of the population. Further, knowledge of the
descriptive profile provides practitioners with the opportunity to make informed decisions about
resource and service delivery to off-campus students with sensitivity to the demographic makeup of the group, rather than operating with a general premise of the group. Furthermore, often
university resources (e.g., staff, budget) dedicated to supporting off-campus students and
commuters are limited, while university support for their residential peers is ample (Dugan et al.,
2008). With recognition of the sheer size of the off-campus population at MSU (N = 13,432),
town-gown administrators can use empirical data gained in the current study to substantiate their
pursuit of additional resource allocation dedicated to supporting off-campus students.
The current study also provided an initial exploration of off-campus students’ sense of
belonging to the local community where they reside. Results of the current study showed
demographic characteristics, except for sex, were not significant contributors to sense of
belonging. For town-gown practitioners, this revelation suggests regardless of demographic
differences or similarities, off-campus students are starting at a similar baseline in terms of their

106

sense of belonging to the local community. Thus, efforts to cultivate belonging should focus on
interactions, activities, and offerings happening in real-time within the local community, instead
of nuanced approaches based on differing demographics. The one caveat was the significance of
the sex variable, which indicated off-campus male students had less sense of belonging than offcampus female students. For practitioners, this difference between males and females suggests
additional effort and interactions with males is warranted to help raise their sense of belonging to
the community.
In support of the emphasis on interactions and activities, the neighborhood perceptions
construct emerged as a significant positive predictor of sense of belonging in off-campus
students, while the neighborhood relations (e.g., social interactions) did not. For practitioners,
this finding underscores the importance of outreach efforts dedicating to influencing positive
perceptions of the local community, such as emphasizing community norms around friendliness,
trust and respect among neighbors, and feelings of responsibility for one’s actions and property.
In contrast, the non-significance of neighborhood interactions suggests traditional town-gown
efforts to encourage neighborly interactions, such as casual greetings or helping with a favor
(Hintz, 2011; McLaughlin, 2011), are not yielding the anticipated gains in sense of belonging.
Town-gown administrators should re-evaluate their outreach efforts and potentially invest more
attention into cultivating positive perceptions instead of championing neighborly interactions
within the environment. That said, participation in positive behaviors in the local community
was also a significant predictor of sense of belonging in off-campus students. Town-gown
administrators should continue to encourage off-campus students to attend community and
neighborhood events and to utilize community services and resources as these engagements are
positively associated with higher sense of belonging. These efforts should include promoting

107

awareness of opportunities and sponsoring neighborhood events designed to engage off-campus
students.
To shed light on student behaviors at the center of town-gown tensions, the current study
also looked at the influence of various factors on off-campus students’ participation in positive
and negative behaviors in the local community. The influence of peers emerged as a significant
predictor in both positive and negative behavior participation. Having higher numbers of
roommates was associated with greater participation in positive and negative behaviors,
suggesting having more peers in your residential unit provided access and awareness of
opportunities for both types of behavior participation. Knowing the influence of roommates,
town-gown practitioners should focus outreach on residential units with higher numbers of
occupants to address negative behaviors, while efforts to encourage positive behaviors may be
best tailored to students living alone or with fewer roommates who appear to have less exposure
to opportunities. Similarly, the neighborhood relations construct was also a positive predictor of
both positive and negative behaviors, suggesting increasing social interactions among neighbors
was associated with participation in both behavior types. As with roommates, there is a
likelihood increasing interactions with neighbors also exposes students to increased opportunities
for both positive and negative behaviors. For town-gown practitioners, this insight means
acknowledging efforts to promote neighborly interactions are both helping to increase desired
positive behaviors in the community, while also simultaneously increasing the undesired
negative behaviors. Town-gown administrators should think deliberately about ways to refine
and target outreach efforts to lessen the potential role they may be playing in cultivating the
negative behaviors they aim to address.

108

Time spent living off-campus and on-campus also emerged as a significant predictor of
behavioral participation. Increasing time spent living off-campus was associated with greater
participation in positive and negative behaviors, suggesting a greater familiarity of opportunity
and access. Greater time living on-campus before moving to the local community emerged as a
positive predictor of negative behavior, suggesting off-campus students free from residence hall
rules are seeking opportunities to engage in negative behaviors (e.g., attend parties, consume
alcohol) and may not be familiar with local community expectations, such as noise and party
ordinances. For town-gown practitioners, these findings underscore a value in focusing tailored
attention and resources on students who are newer to off-campus living and those who have
previously lived in residence halls. Newer student residents to the local community would
benefit from exposure to resources outlining positive behavior opportunities, while also
clarifying behavioral expectations around municipal ordinances and neighborhood norms.
Interestingly, increasing perception of problems in the local environment was a positive
predictor of positive and negative behaviors. As off-campus students had an increasing sense of
community problems, they seemed to be motivated to access and utilize community services and
resources (e.g., positive behaviors) to address their concerns, while simultaneously be validated
to participate in negative behaviors, such as attending or hosting parties and consuming alcohol.
This finding presents a challenge because town-gown practitioners want off-campus students to
feel empowered to seek solutions to perceived problems but do not want students to feel like the
presence of the problems (e.g., presence of noise, litter, unkempt properties) makes participation
in negative behaviors less significant or problematic to the local community. Town-gown
practitioners might be best to consider efforts to further support civic agency while also raising
awareness of the harm caused by negative behaviors in the community.

109

Residential density proved to be a significant predictor in positive and negative
behaviors. For positive behaviors, all levels of density associated with living in a house were
significant, while none of the apartment levels were significant, suggesting house dwellers
participate in positive behaviors more than their apartment peers. In recognition of this finding,
town-gown practitioners should refine outreach efforts to better illuminate opportunities for
positive behavior participation among apartment dwellers. For negative behaviors, areas of high
house density emerged as the sole positive predictor of negative behavior participation. For
town-gown practitioners, this finding confirms geographic areas with a high density of houses
and many off-campus residents are centers of negative behaviors. Efforts to address negative
behaviors should continue to be wide-spread, but extra effort and strategies specific to the highdensity residential areas are warranted.
Lastly, sense of belonging emerged as a positive predictor of positive behaviors but not
of negative behaviors. For town-gown administrators, this finding confirms efforts to promote
sense of belonging to the local community do influence greater participation in positive
behaviors, while having no effect in negative behavior participation. Town-gown practitioners
are faced with the notion their efforts to cultivate belonging are not translating into reduction in
negative behaviors, but are contributing greater positive behavior participation in the local
community instead. This reality suggests efforts designed around promoting belonging do have
merit and will translate into greater participation in positive behaviors from off-campus students.
As sense of belonging increases among more off-campus students, perhaps greater participation
in positive behaviors will influence the overall wellbeing and harmony of the local community.
At the same time, town-gown practitioners must recognize other factors (i.e., environmental
factors, density, neighborhood relations) are more significant influencers of negative behavior

110

participation and subsequent outreach efforts and resources should be adapted and tailored to
address factors that make a difference in negative behavior participation.
Recommendations for Future Research
The current study focused on off-campus students as a subpopulation of commuter
students and central actors in the town-gown context. Despite initial insight attained through this
study, additional research on off-campus students is warranted. Future research should delve
into understanding off-campus students as a subpopulation of college students and to identify
their unique student experience, challenges, and needs for support within the collegiate context.
Additionally, the current study focused on undergraduate students specifically because of their
role in student behavior problems in the town-gown context. There are no doubt graduate
students who are also part of the off-campus population and their experience should also be
explored and included in what is known about off-campus students. Future use of qualitative
methodology may be ideal to explore the experience of off-campus students and bring their voice
to the town-gown context specifically and higher education, more broadly.
Density was a new variable used in the current study, which proved to be a significant
one. Initial findings of influence associated with density should be examined more fully to
confirm the directionality and power of influence found in the current study. Additionally,
density of houses and number of roommates emerged as significant predictors of behaviors.
Considering this finding, future research should differentiate between single rental houses likely
to have fewer roommates and larger group houses, like fraternity and sorority houses, with many
student residents. There may also be merit in further distinguishing different apartment types
(e.g., large, multiple-building complex, small single building) and their corresponding densities.
Relatedly, future exploration of dwelling unit type (e.g., apartment, house) would illuminate how

111

physical design and structure of dwelling type influences feelings of belonging regardless of
density. Another important consideration associated with density was the notion of exploring
density as a measure of individuals in a geographic area (as was the case in the current study)
and the idea of density of relationships, or social network, in a geographic area. Further
exploration of both density types would provide insight on how peer networks and physical
proximity to peers compare in their influence on off-campus students in the local community.
One of the two types may prove to be a more valuable indicator of the experience of off-campus
students in the town-gown context.
The construct for positive and negative behavior participation was a new component
designed specifically for the current study. Although a reliable measure in the current study, the
new behavior construct should be further evaluated to confirm its validity and explore its utility
across different samples and contexts. There is an opportunity to use pilot testing to explore
expanding the lists of behaviors included in the construct. The positive behaviors construct
could grow to include a balance of items measuring positive event or initiative participation (e.g.,
attend a neighborhood event; use city of services) and positive neighboring behaviors (e.g., yard
maintenance, care and respect for neighbors). For negative behaviors, consideration should be
given to delineating between lesser negative behaviors (e.g., attending a party, consuming
alcohol) and more serious negative behaviors where there may be a judicial or municipal
consequence possible (e.g., noise or alcohol violation, fight). Exploratory factor analysis could
help to discern naturally occurring patterns within the list of negative behaviors. Efforts to
develop a more comprehensive but refined set of behaviors for town-gown research would
strengthen future research utilizing the behaviors construct and ultimately better inform towngown practice.

112

The sense of belonging off-campus students felt to the local community was the central
component of the current study. This study was an initial effort at exploring sense of belonging
to a community external to the university itself and beyond campus boundaries. The intent in
studying an external community was to acknowledge students are likely to experience feelings of
belonging to multiple communities that together influence their lived experience in college. Just
as there is credence in recognizing the multiple identities students possess as they interact in the
college context and world at large, there may be value in also understanding how their feelings
of belonging across various communities influence their student experience. Further exploration
seems worthwhile given the known importance sense of belonging has in eliciting positive
student outcomes in higher education. Future research should continue to explore the sense of
belonging of college students across their co-existing communities of support and influence.
Lastly, GIS mapping was a tool employed in the current study that would be a powerful
addition to future studies in higher education across a myriad of dimensions and topics. GIS
mapping provides the opportunity to visualize, analyze, and interpret data of all kinds to
understand relationships and patterns. Demographic data from the current study could be further
mapped to reveal patterns and trends among off-campus students. For example, mapping
students by age, year in school, or sex might reveal patterns by housing type (e.g., house,
apartment), density level, or distance from campus that exist. The same mapping in subsequent
years could reveal patterns of transience or movement among off-campus students. This type of
data-driven knowledge would be extremely valuable to town-gown practitioners and policy
makers as they make resource decisions (e.g., materials, staffing) about outreach efforts and
strategies for working with off-campus students and town-gown relations. GIS mapping
employed in higher education research and practice could aid in decision making for program

113

and service offerings and implementation, improve organizational efficiencies around staffing
demands and scheduling, and promote data-driven fiscal and human resource allocation for
research projects and administrative initiatives.
Summary
The current chapter included a restatement of the purpose of this study, a discussion of
the findings associated with each research question, a consideration of implications for research
and practice, and suggestions for future research. The current study emerged from the need to
address student behavior issues in the town-gown community and the practical need to know
more about off-campus students as central actors in these behaviors. This study validates the
need to further explore off-campus students as a unique cohort within the college student
population as the sample revealed off-campus students were distinct from traditional perceptions
of commuter students. Analyses confirmed town-gown efforts to promote perceptions of
community belonging and to encourage participation in positive behaviors yield increases in
sense of belonging to the local community. Findings suggested town-gown strategies to
influence off-campus student behaviors should differ based on residential density of houses and
apartments. Peer relations were found to be instrumental in encouraging participation in both
positive and negative behaviors, confirming town-gown strategies aimed at reducing problem
behaviors should continue to focus attention on social relations and promoting positive behaviors
as normative in the local community. The factors and characteristics identified to predict sense
of belonging and behavior participation should be used to augment existing town-gown strategies
to support off-campus students, while also guiding the careful design and purposeful
implementation of new efforts to address student behavior issues in the local community. The
current findings provide an empirical baseline from which to move forward in policy and

114

practice. Lastly, this study has demonstrated the importance of focused attention on supporting
off-campus students and advocates for their inclusion in future town-gown research. This
inclusion is essential if universities and communities are serious about wanting a more
harmonious and supportive town-gown relationship for the betterment of both entities.

115

APPENDICES

116

APPENDIX A
Map 1

Figure 3. Map 1: Off-Campus Students in City of East Lansing ZIP Code
117

APPENDIX B
Map 2

Figure 4. Map 2: Off-Campus Students in East Lansing ZIP Code with Grid Overlay

118

APPENDIX C
Map 3

Figure 5. Map 3: Density Categories Based on Off-Campus Students in East Lansing ZIP
Code
119

APPENDIX D
Map 4

Figure 6. Map 4: Density Categories and Off-Campus Student Data in East Lansing ZIP
Code
120

APPENDIX E
Participant Consent Form for Research Study
Study title: Examining Sense of Belonging and Student Behaviors in a Town-Gown Context
Researcher: Erin Carter, doctoral candidate in Higher, Adult, and Lifelong Education (HALE)
Department and Institution: Educational Administration, Michigan State University
Purpose of Research:
The purpose of this research project is to learn about your sense of belonging to the off-campus
community you reside in during your time as a student at Michigan State University. I am
interested in understanding which factors contribute to your sense of belonging to the
community, such as the neighborhood environment and your interactions with community
members. I am also interested in examining how your sense of belonging may influence your
participation in a range for behaviors within the community from social activities to civic
responsibilities.
You will be asked to complete a survey with questions regarding your experience and actions
living off-campus in the City of East Lansing while a student at Michigan State University.
Your survey answers will be used for further statistical analysis by me only. You must be at
least 18 years old to participate in this research.
Potential Benefits: There is a possibility you will not benefit from participating in this research.
However, all responders to the survey are offered an opportunity to enter a drawing for one
chance to win $100 toward their rent expense. If you choose to enter this drawing, your contact
information will be collected by a separate system and will not be connected in any way to your
survey responses. You are not required to finish the survey to enter the drawing.
Potential Risks:
There are no foreseeable risks associated with participation in this study. However, if you wish,
you may, at any time and without penalty, elect not to answer a question on the survey or you
may discontinue participating without fear of any harm or repercussions.
Confidentiality and Privacy:
The data for this project are being collected by survey using a software called Qualtrics. The
Qualtrics software disassociates identifiable student information from survey responses so the
researchers will NOT be able to link your survey data to you. Survey responses from all
participants will be pooled and reported collectively to further aid in maintaining confidentiality.
Voluntary Participation:
Participation in this research project is completely voluntary. You have the right to say no. You
may change your mind at any time and withdraw. You may choose not to answer specific
questions or to stop participating at any time. Whether you choose to participate or not will have
no effect on you or any potential benefits from this research.
121

Request for Additional Information:
If you have any questions about this research, please contact Erin Carter, the doctoral student
conducting this research study at carte221@msu.edu. If you have concerns regarding your rights
as a study participant, or are dissatisfied at any time with any aspect of this study, you may
contact – anonymously, if you wish, Dr. Matthew Wawrzynski, 429 Erickson Hall, Michigan
State University, (517) 355-6617.
Clicking the “I Agree” button and beginning the survey, indicates your voluntary agreement to
participate in this study.

122

APPENDIX F
Sense of Belonging Survey Instrument
Instructions: This survey is designed to learn more about your sense of belonging to the local
community you reside in during college. Please answer each of the questions on the survey.
Individual responses will be anonymous. Thank you for your responses.
Sense of Belonging Measures
Instructions: Complete the following questions about your sense of belonging to the City of East
Lansing community
1= Strongly Disagree 2= Disagree 3= Neutral 4= Agree 5= Strongly Agree
1

I feel that I am a member of the East Lansing community.

1 2 3 4 5

2

I see myself as part of the East Lansing community.

1 2 3 4 5

3

I feel a sense of belonging to the City of East Lansing.

1 2 3 4 5

4

I feel a sense of belonging to Michigan State University.

1 2 3 4 5

5

In thinking about my experience as both an MSU student and a resident of East
Lansing, I feel my sense of belonging is:
stronger to East
Equal between East
Lansing than MSU Lansing and MSU

stronger to MSU than to neither East
East Lansing
Lansing or MSU

Factors Influencing Sense of Belonging
I. Perception of Belonging
Instructions: Complete the following questions about your perceptions of belonging to your
specific residential neighborhood using the following scale:
1= Strongly Disagree 2= Disagree 3= Neutral 4= Agree 5= Strongly Agree
6

There are people I can rely on among my neighbors.

1 2 3 4 5

7

People trust each other in my neighborhood.

1 2 3 4 5

8

I feel I belong in my neighborhood.

1 2 3 4 5

9

I care about what my neighbors think of my actions (e.g., if I
take care of my property, how I act, etc.)?

1 2 3 4 5

10

I feel close to some of my neighbors.

1 2 3 4 5

11

People in my neighborhood are usually warm and friendly.

1 2 3 4 5

12

We look out for one another in this neighborhood.

1 2 3 4 5

123

II. Perception of the Environment
Instructions: In thinking about your neighborhood of residence, select the level of significance
you would give each potential problem in your neighborhood:
1= Not a problem
2= Minor problem
3= Moderate problem
4= Major problem
13 Condition of houses

1 2 3 4

14 Traffic

1 2 3 4

15 Street parking

1 2 3 4

16 Crime

1 2 3 4

17 Litter

1 2 3 4

18 Noisy neighbors

1 2 3 4

19 Unkempt lawns

1 2 3 4

20 Garbage collection

1 2 3 4

21 Recycling collection

1 2 3 4

22 Police services
a) do you see this as enough, not enough, too much?

1 2 3 4

23 Parking services (enforcement)
a) do you see this as enough, not enough, too much?

1 2 3 4

III. Neighborhood Relations
Instructions: Answer the following questions regarding your interactions with your neighbors.
1= Very Unlikely 2= Unlikely
3= Not Sure
4= Likely
5= Very Likely
24

How likely is it that I would ask a neighbor to borrow something
like a hammer (if they had one and I need one)?

1 2 3 4 5

25

How likely is it that I would ask a neighbor to borrow something
like money?

1 2 3 4 5

26

How likely is it that a neighbor would ask me to borrow
something, like a hammer (if I had one and they needed one)?

1 2 3 4 5

27

How likely is it that a neighbor would ask me to borrow
something, like money?

1 2 3 4 5

28

How likely is it that I would get help from a neighbor? (e.g.,
watch my place if I am away, feed my pet, collect my mail)?

1 2 3 4 5

29

How likely is it that I would help a neighbor (e.g., watching their
place if they are away)?

1 2 3 4 5

124

Instructions: Answer the following regarding your interactions with your neighbors.
1= Very Seldom/Never 2= Seldom (once every 3 months)
3= Sometimes (once per month)
4=Often (once per week)
5= Very Often (daily)
30

How often do I greet my neighbors when I see them?

1 2 3 4 5

31

How often do I casually visit with neighbors, either going over to
their place or them coming to mine?

1 2 3 4 5

32

How often do I participate in neighborhood activities (e.g.,
cookouts, parties)?

1 2 3 4 5

33

How often do I talk to neighbors who are students?

1 2 3 4 5

34

How often do I talk to neighbors who are non-students?

1 2 3 4 5

35

How often do people in this neighborhood do favors for each
other?

1 2 3 4 5

Participation in Behaviors
Instructions: Indicate the frequency in which you have participated in each of the following
activities or actions in the past six months.
1= Very Seldom/Never 2= Seldom (once every 3 months) 3= Sometimes (once per
month)
4=Often (once per week)
5= Very Often (daily)
36

When seasonally appropriate, how often do I do yard maintenance,
like mowing, raking, or shoveling?

1 2 3 4 5

37

How often do I use trash and recycling services at my residence?

1 2 3 4 5

38

How often do I attend a party in my neighborhood?

1 2 3 4 5

39

How often do I host a party at my residence?

1 2 3 4 5

40

How often do I consume alcohol?

1 2 3 4 5

41

How often do I receive an alcohol violation from police?

1 2 3 4 5

42

How often is loud noise coming from my residence (e.g. music)?

1 2 3 4 5

43

How often have I received a noise violation ticket at my
residence?

1 2 3 4 5

44

How often do I get into a fight with people in my neighborhood?

1 2 3 4 5

45

How often do I pick up litter or trash in my yard?

1 2 3 4 5

46

How often do I participate in community-wide activities (e.g.,
Farmer’s Market, Taste of East Lansing, live music)?

1 2 3 4 5

47

How often do I use city resources or services (e.g., public library,
community center, parks, recreation facilities)?

1 2 3 4 5

125

Demographic Information
Instructions: Answer the following questions about you and your place of living.
During spring semester 2016:
48 What type of property do House
I live in?

Apartment Housing
Greek
Cooperativ House
e

49 Who do I live with?
Alone
- sub question: what
percentage of the people
who live with me are also
affiliated with MSU?

With
With
Partner/sign Other
roommate parent/gua ificant other (please list)
s (list #)
rdian

50 a) What percentage of
people who live near
me are also affiliated
with MSU?
b) What percentage of
people are not
affiliated with MSU?

Other
(please list)

Enter percentages for both parts.

Think about your MSU college experience,
51 Prior to living offcampus, how many
semesters did I live oncampus in a residence
hall or university
apartment?

0

1

2

3

4

52 How many semesters
have I lived off-campus,
including the current
semester?

1

2

3

4

5

Demographics - about you
53 What is my age?

Enter a number in years

54 What is my gender?

Male

Female

Transgender

55 What is my year in
school?

First
year

Second
year

Third year

56 What is my citizenship
status?

Domestic student
(U.S. Citizen)

126

Fourth Year Other
(please list)

International student

During spring semester 2016:
57 Which best describes my _____ Black or African American
racial/ethnic background?
Please check all the
_____ White (Not Hispanic/Not Latino)
apply.
_____ Hispanic or Latino
_____ American Indian
_____ Asian
_____ Native Hawaiian or Pacific Islander
_____ Multiracial (Having parents of more than one race)
_____ Member of race not listed above: ________________
Survey End - Optional drawing entry
58 Do you wish to enter the drawing for a chance
to win $100 for use toward your rent?

127

Yesbranch to
drawing

No - branch to thank you
for completing the study

APPENDIX G
Construct Development and Alpha Reliability for Items Used in Study
Table G1
Construct Development and Alpha Reliability for Items Used in Study
Sense of Belonging (SoBT)
Mean

SD

Alpha
if item
deleted

Likert scale: 1 = Strongly disagree to 5 = Strongly Agree
I feel that I am a member of the East Lansing community.
I see myself as part of the East Lansing community
I feel a sense of belonging to the City of East Lansing
Cronbach’s alpha reliability

3.77
.942
3.77
.944
3.74
.957
.907

Environmental Factors (Factors)

Mean

SD

Alpha
if item
deleted

Likert scale: 1 = Not a Problem to 4 = Major Problem
Condition of houses
Traffic
Street parking
Crime
Litter
Noisy neighbors
Unkempt lawns
Garbage collection
Recycling collection
Police services
Parking enforcement services

1.93
1.93
2.63
2.03
2.21
2.15
1.49
1.46
1.95
1.49
2.20

.865
.889
1.160
.896
.889
.942
.706
.726
1.117
.795
1.176

.774
.775
.782
.771
.767
.780
.771
.772
.796
.770
.786

Mean

SD

Alpha
if item
deleted

3.18
3.33
3.39
3.37

1.112
.918
.953
1.030

.843
.858
.853
.881

2.86
1.189
3.54
.867
3.12
.961
.874

.848
.860
.845

Cronbach’s alpha reliability
Neighborhood Perceptions (Perceptions)
Likert scale: 1 = Strongly Disagree to 5 = Strongly Agree
There are people I can rely on among my neighbors.
People trust each other in my neighborhood.
I feel I belong in my neighborhood.
I care about what my neighbors think of my actions (e.g., if
I take care of my property, how I act, etc.)?
I feel close to some of my neighbors.
People in my neighborhood are usually warm and friendly.
We look out for one another in this neighborhood.
Cronbach’s alpha reliability
128

.850
.839
.909

.793

Table G1 (cont’d)
Neighborhood Relations (Relations)
Likert scale: 1 = Very Unlikely to 5 = Very Likely
How likely is it that I would ask a neighbor to borrow
something like a hammer (if they had one and I need one)?
How likely is it that I would ask a neighbor to borrow
something like money?
How likely is it that a neighbor would ask me to borrow
something, like a hammer (if I had one and they needed
one)?
How likely is it that a neighbor would ask me to borrow
something, like money?
How likely is it that I would get help from a neighbor?
(e.g., watch my place if I am away, collect mail)?
How likely is it that I would help a neighbor (e.g.,
watching their place if they are away)?
Likert Scale: 1 = Very Seldom/Never to 5 = Very Often (daily)
How often do I greet my neighbors when I see them?
How often do I casually visit with neighbors, either going
over to their place or them coming to mine?
How often do I participate in neighborhood activities (e.g.,
cookouts, parties)?
How often do I talk to neighbors who are students?
How often do I talk to neighbors who are non-students?
How often do people in this neighborhood do favors for
each other?
Cronbach’s alpha reliability
Positive Behaviors (PosBehav)
Likert Scale: 1 = Very Seldom/Never to 5 = Very Often (daily)
When seasonally appropriate, how often do I do yard
maintenance, like mowing, raking, or shoveling?
How often do I use trash and recycling services at my
residence?
How often do I pick up litter or trash in my yard?
How often do I participate in community-wide activities
(e.g., Farmer’s Market, Taste of East Lansing, live music)?
How often do I use city resources or services (e.g., public
library, community center, parks, recreation facilities)?
Cronbach’s alpha reliability
129

Mean

SD

Alpha
if item
deleted

3.05

1.353

.896

1.30

.692

.909

2.89

1.301

.896

1.42

.793

.909

2.56

1.327

.898

3.25

1.284

.903

3.13
1.98

1.280
1.230

.901
.896

1.86

1.108

.901

2.85
1.66
2.02

1.346
.991
1.070

.899
.909
.896

Mean

SD

Alpha
if item
deleted

2.10

1.247

.528

3.86

.926

.626

2.31
1.98

1.235
.976

.519
.523

2.54

1.184

.572

.909

.611

Table G1 (cont’d)
Negative Behaviors (NegBehav)
Likert Scale: 1 = Very Seldom/Never to 5 = Very Often (daily)
How often do I attend a party in my neighborhood?
How often do I host a party at my residence?
How often do I consume alcohol?
How often do I receive an alcohol violation from police?
How often is there loud noise coming from my residence
(e.g. music)?
How often have I received a noise violation ticket at my
residence?
How often do I get into a fight with people in my
neighborhood?
Cronbach’s alpha reliability

130

Mean

SD

Alpha
if item
deleted

2.28
2.13
3.40
1.16
2.25

1.163
1.026
1.097
.602
1.151

.672
.670
.733
.712
.709

1.21

.624

.703

1.18

.558

.716

.735

APPENDIX H
Correlation Matrix of Predictor and Criterion Variables
Table H1
Correlation Matrix of Predictor and Criterion Variables
Measure
1. Age

1
1

2

3

4

5

6

7

2. Sex

.124**

1

3. Race1

-.078

-.068

1

4. Race2

.062

.028

-.808**

1

5. Race3

.041

.074

-.522**

-.080

1

6. YrSchool

.717**

.045

.023

.053

-.115**

1

.034

-.035

-.007

.158**

1

7. LiveOn

-.033

-.127

**

8

9

10

11

12

13

14

8. LiveOff

.499

**

.019

.043

-.024

-.038

.617**

-.082*

1

9. RoomLog

-.133**

.017

.103*

-.063

-.083*

-.019

.116**

.057

1

10. AptHigh

-.157**

-.100*

.007

.025

-.049

-.196**

.045

-.171**

-.039

11. AptMed

.014

.017

-.092*

.035

.106*

.006

.005

-.062

-.234** -.353**

12. AptLow

.093*

-.039

.000

-.030

.043

.078

-.071

.087*

-.248** -.202** -.162**

-.005

.028

.014

-.068

.021

-.042

.068

.210**

-.198** -.159**

-.274** -.220** -.126** -.123**

13. ResHigh

-.032

15

16

17

18

19

20

1
1
1
-.091*

1

14. ResMed

.087*

.069

.115**

-.100*

-.050

.119**

.086*

.105*

.275**

15. ResLow

.050

.073

-.036

.037

.007

.049

-.059

.067

.080

16. SoBT

-.073

-.083*

.029

-.015

-.028

-.040

.057

.025

.066

.026

-.122**

-.010

.073

.040

.020

1

17. Perception

-.061

.039

-.011

-.021

.049

-.049

.082*

.061

.073

.028

-.144**

-.030

.107**

.085*

-.011

.421**

1

18. Factors

-.015

-.046

.015

-.059

.060

.008

-.030

.091*

.108**

-.011

.002

-.045

.161**

-.023

-.052

.026

-.001

**

**

-.048

.172

**

*

-.013

.286

.066

1

-.137** -.237**

-.053

.218**

.226**

.101*

.218**

.310**

.139**

.398**

1

-.138**

-.086*

.257**

.119**

-.052

.141**

.330**

.207**

.495**

.370**

19. Relations

-.061

.067

-.009

-.031

.060

-.059

.036

.069

20. PosBehav

.056

.022

.091*

-.081*

-.036

.089*

.025

.164**

.153

.240**

21.NegBehav

-.104*

.174**

.057

-.082*

.022

-.059

.054

.052

.305**

21

1

-.291** -.233** -.134** -.131** -.181**

-.016
-.026

Note. * p < 0.05, ** p < 0.01 level (2-tailed)

131

-.136

.099

1

**

.707

**

1

1

APPENDIX I
Collinearity Statistics for Final Regression Models
Table I1
Research Question One: Collinearity Statistics for Final Hierarchical Regression Model of
Background Characteristics and Density Levels on Sense of Belonginga (n=589)
Model 2

Unstandardized
Coefficients
B

Standardized
Coefficients

Std. Error

b

Collinearity
Diagnostics
t

Sig.

Tolerance

VIF

Step 1:
Age
Sex
Race2
Race3
YrSchool
LiveOn
LiveOff
RoomLog

-.134
-.365
.019
-.096
-.190
.127
.178
-.018

.139
.229
.347
.514
.211
.087
.112
.526

-.061
-.067
.002
-.008
-.064
.065
.087
-.002

-.969
-1.596
.054
-.186
-.904
1.457
1.600
-.034

.333
.111
.957
.852
.366
.146
.110
.973

.432
.957
.978
.936
.339
.857
.568
.936

2.316
1.045
1.022
1.068
2.952
1.167
1.760
1.068

Step 2:
AptMed
AptLow
ResHigh
ResMed
ResLow

-.592
-.072
.563
.239
.139

.306
.431
.436
.361
.337

-.094
-.008
.059
.032
.020

-1.938
-.168
1.291
.663
.413

.053
.867
.197
.508
.680

.709
.789
.801
.703
.743

1.410
1.268
1.249
1.423
1.345

Note.

a

Results presented are from the second and final step of the analysis.

132

Table I2
Research Question Two: Collinearity Statistics for Final Hierarchical Regression Model of
Demographic, Density, Factors, Perceptions, Relations, and Behavioral variable on Sense of
Belonginga (n=589)
Model 6
Unstandardized Standardized
Collinearity
Coefficients
Coefficients
Diagnostics
B
Std.
t
Sig.
Tolerance
VIF
b
Error
Step 1
Age
-.091
.127
-.041
-.714
.475
.426
2.346
Sex
-.465
.215
-.086
-2.168
.031
.895
1.117
Race2
.036
.318
.004
.113
.910
.959
1.043
Race3
-.376
.473
-.031
-.795
.427
.910
1.099
YrSchool
-.037
.193
-.013
-.193
.847
.331
3.018
LiveOn
.029
.080
.015
.367
.714
.830
1.205
LiveOff
.025
.103
.012
.239
.811
.547
1.830
RoomLog
.057
.493
.005
.116
.908
.426
1.400
Step 2
AptMed
AptLow
ResHigh
ResMed
ResLow

-.222
.051
.042
-.077
.098

.282
.395
.413
.337
.312

-.035
.005
.004
-.010
.014

-.788
.130
.103
-.228
.314

.431
.897
.918
.820
.754

.697
.786
.745
.672
.720

1.434
1.272
1.342
1.487
1.388

Step 3
Factors

.007

.017

.015

.378

.706

.906

1.104

Step 4
Perceptions

.208

.026

.425

7.927

.000

.486

2.059

Step 5
Relations

-.014

.015

-.052

-.897

.370

.408

2.449

.695
.614

1.438
1.627

Step 6
PosBehav
.076
.033
.102
2.265
.024
NegBehav
-.006
.031
-.010
-.204
.839
a
Note. Results presented are from the sixth and final step of the analysis.

133

Table I3
Research Question Three: Collinearity Statistics for Final Hierarchical Regression Model of
Background, Density, and Sense of Belonging on Positive Behavior Participationa (n=589)
Model 6

Unstandardized
Coefficients
B

Std. Error

Step 1:
Age
Sex
Race2
Race3
YrSchool
LiveOn
LiveOff
RoomLog

.125
-.091
-.562
-.459
.064
.006
.131
1.046

.161
.267
.401
.598
.245
.101
.131
.613

Step 2:
AptMed
AptLow
ResHigh
ResMed
ResLow

-.444
.309
2.097
1.955
1.407

Step 3
Factors

Standardized
Coefficients
b

Collinearity
Diagnostics
t

Sig.

Tolerance

VIF

.042
-.013
-.050
-.028
.016
.002
.048
.070

.777
-.343
-1.401
-.767
.263
.059
1.004
1.706

.437
.732
.162
.443
.793
.953
.316
.089

.428
.928
.964
.910
.331
.832
.548
.739

2.336
1.078
1.037
1.099
3.017
1.202
1.824
1.353

.356
.499
.512
.418
.390

-.053
.025
.164
.197
.148

-1.249
.619
4.096
4.677
3.610

.212
.536
.000
.000
.000

.699
.787
.775
.698
.741

1.431
1.271
1.290
1.432
1.350

.054

.022

.091

2.509

.012

.938

1.066

Step 4
Perceptions

.017

.035

.026

.493

.622

.438

2.284

Step 5
Relations

.102

.018

.287

5.628

.000

.476

2.099

Step 6
SoBT

.119

.053

.089

2.263

.024

.804

1.243

Note.

a

Results presented are from the sixth and final step of the analysis.

134

Table I4
Research Question Three: Collinearity Statistics for Final Hierarchical Regression Model of
Background, Density, and Sense of Belonging on Negative Behavior Participationa (n=589)
Model 6

Unstandardized
Coefficients

Standardized
Coefficients

Collinearity
Diagnostics

B

Std. Error

b

t

Sig.

Tolerance

VIF

Step 1:
Age
Sex
Race2
Race3
YrSchool
LiveOn
LiveOff
RoomLog

-.240
1.428
-.555
.035
-.111
.154
.135
2.897

.174
.289
.434
.647
.265
.110
.141
.664

-.070
.171
-.043
.002
-.024
.051
.043
.169

-1.382
4.947
-1.278
.054
-.419
1.398
.956
4.365

.167
.000
.202
.957
.675
.163
.339
.000

.428
.928
.964
.910
.331
.832
.548
.739

2.336
1.078
1.037
1.099
3.017
1.202
1.824
1.353

Step 2:
AptMed
AptLow
ResHigh
ResMed
ResLow

-.301
-.040
1.732
.207
-.523

.385
.540
.554
.453
.422

-.031
-.003
.118
.018
-.048

-.782
-.073
3.124
.458
-1.240

.435
.942
.002
.647
.215

.699
.787
.775
.698
.741

1.431
1.271
1.290
1.432
1.350

Step 3
Factors

.097

.023

.141

4.115

.000

.938

1.066

Step 4
Perceptions

-.017

.038

-.023

-.457

.648

.438

2.284

Step 5
Relations

.172

.020

.424

8.805

.000

.476

2.099

Step 6
SoBT

.009

.057

.006

.154

.878

.804

1.243

Note.

a

Results presented are from the sixth and final step of the analysis.

135

APPENDIX J
Sample Variance Tests for Sense of Belonging
Table J1
Sample T-test of Sense of Belonging as a Function of Sex (N = 589)
Female
Male
n = 376
n = 213
Construct
Mean
SD
Mean
SD
SoBT

11.44

2.58

11.00

2.65

df

t

587

.045*

Note. * p < .05, ** p < .01
Table J2
Means, Standard Deviations, One-way Analysis of Variance on Sense of Belonging as a Function of Race (N = 589)
White
People of color
International
n = 495
n = 65
n = 29
F
Construct

Mean

SD

Mean

SD

Mean

SD

(3, 586)

h2

SoBT

11.32

2.57

11.17

3.11

10.97

2.096

.313

.00

Note. * p < .05, ** p < .01

Table J3
Means, Standard Deviations, One-way Analysis of Variance on Sense of Belonging as a Function of Year in School (N = 589)
First
Second
Third
Fourth
Fifth
Other
n=3
n = 92
n = 224
n = 228
n = 35
n=7
F
Construct Mean SD Mean SD Mean SD
Mean
SD
Mean
SD
Mean
SD (5, 583) h2
SoBT
11.67 2.89 11.29 2.41 11.39 2.40 11.27
2.76
10.69
3.29
11.00
3.27
.47
.00
Note. * p < .05, ** p < .01
136

Table J4
Means, Standard Deviations, One-way Analysis of Variance on Sense of Belonging as a Function of Semesters Lived OnCampus (N = 589)
Zero
n = 88
Construct Mean
SD
SoBT
10.51
2.69
Note. * p < .05, ** p < .01

One
n = 45
Mean
12.07

SD
2.21

Two
n = 276
Mean
11.39

Three
n = 15

SD
2.53

Mean
10.60

Four
n = 165

SD
2.06

Mean
11.36

SD
2.77

F
(4, 584)
3.40**

h2
.02

Table J5
Means, Standard Deviations, One-way Analysis of Variance on Sense of Belonging as a Function of Semesters Lived OffCampus (N = 589)
One
Two
Three
Four
Five
n = 12

n = 211

n = 21

n = 186

n = 159

F

SD

Mean

SD

Mean

SD

Mean

SD

Mean

SD

(4, 584)

h2

SoBT
12.08
1.56
Note. * p < .05, ** p < .01

11.20

2.68

10.86

2.99

11.17

2.70

11.51

2.42

.853

.01

Construct

Mean

Table J6
Means, Standard Deviations, One-way Analysis of Variance on Sense of Belonging as a Function of Density (N = 589)
AptHigh
AptMed
AptLow
ResHigh
ResMed
ResLow
Fa
Construct Mean
SD
Mean SD
Mean
SD Mean
SD
Mean
SD
Mean SD (5, 583) h2
SoBT
11.38 2.64
10.68 2.97 11.20 2.63 11.92 2.16 11.53 2.52 11.40 2.18
2.08
.02
Note. * p < .05, ** p < .01. a Welch’s robust test of equality of means used due to violation of assumption of homogeneity.
n = 180

n = 130

n = 50

n = 48

137

n = 86

n =95

APPENDIX K
Hierarchical Regression Analysis for Research Question One
Table K1
Research Question One: Hierarchical Regression of Demographic and Density on Sense of
Belonging (n=589)
Model 1
Model 2
Predictors
Step 1
Age
Sex
Race2
Race3
YrSchool
LiveOn
LiveOff
RoomLog
Step 2
AptMed
AptLow
ResHigh
ResMed
ResLow
R
R2
Adj. R2
∆R2
∆F
Note. * p < .05, ** p < .01, *** p < .001

b

b

-.056
-.066
.000
-.021
-.069
.061
.099
.044

-.056
-.066
.002
-.008
-.064
.065
.087
-.002
-.094
-.008
.059
.032
.020

.148
.022
.008

138

.190
.036
.014
.014
1.702

APPENDIX L
Hierarchical Regression Analysis for Research Question Two
Table L1
Research Question Two: Hierarchical Regression of Demographics, Density, Factors,
Perceptions, Relations, and Behaviors on Sense of Belonging (n=589)
Model 1
Model 2
Model 3
Model 4
Model 5
Model 6
Predictors
Step 1
Age
Sex
Race2
Race3
YrSchool
LiveOn
LiveOff
RoomLog
Step 2
AptMed
AptLow
ResHigh
ResMed
ResLow
Step 3
Factors
Step 4
Perceptions
Step 5
Relations
Step 6
PosBehav
NegBehav

b
-.056
-.066
.000
-.021
-.069
.061
.099
.044

b

b

b

-.061
-.067
.002
-.008
-.064
.065
.087
-.002

-.060
-.067
.003
-.009
-.064
.065
.086
-.003

-.036
-.090*
.000
-.035
-.009
.015
.015
.009

-.036
-.089*
.000
-.034
-.011
.015
.017
.011

-.041
-.086*
.004
-.031
-.013
.015
.012
.005

-.094
-.008
.059
.032
.020

-.094
-.007
.058
.033
.020

-.041
.008
.018
.009
.029

-.041
.008
.020
.010
.030

-.035
.005
.004
-.010
.014

.010

.022

.023

.015

.413**

.432**

.425**

-.028

-.052

R
.148
.190
2
R
.022
.036
2
Adj. R
.008
.014
.014
R2∆
1.702
F∆
Note. * p < .05, ** p < .01, *** p < .001

b

b

.102*
-.010
.190
.036
.013
.000
.061

139

.442
.195***
.174
.159
113.198***

.442
.196***
.173
.000
.260

.450
.203***
.178
.007
2.578

APPENDIX M
Hierarchical Regression Analysis for Research Question Three: Positive Behaviors
Table M1
Research Question Threea: Hierarchical Regression of Demographics, Density, Factors,
Perceptions, Relations, and Sense of Belonging on Positive Behavior Participation (n=589)
Model 1
Model 2
Model 3
Model 4
Model 5
Model 6
Predictors
Step 1
Age
Sex
Race2
Race3
YrSchool
LiveOn
LiveOff
RoomLog
Step 2
AptMed
AptLow
ResHigh
ResMed
ResLow
Step 3
Factors
Step 4
Perceptions
Step 5
Relations
Step 6
SoBT

b

b

b

b

b

b

.032
.018
-.066
-.022
-.022
.015
.147**
.228**

.021
.000
-.054
.001
-.036
.029
.116*
.093*

.025
.005
-.049
-.007
-.035
.032
.105*
.084

.039
-.010
-.051
-.023
-.001
.001
.061
.091*

.039
-.020
-.050
-.031
.015
.004
.049
.071

.042
-.013
-.050
-.028
.016
.002
.048
.070

-.088
.014
.226***
.219***
.141***

-.089*
.017
.214***
.223***
.146***

-.056
.027
.189***
.208***
.152***

-.056
.025
.165***
.198***
.150***

-.053
.025
.164***
.197***
.148***

.098**

.106**

.093**

.091**

.259***

.065

.026

.284***

.287***
.089*

R
.292
.421
.432
.499
.536
R2
.085***
.177***
.186***
.249***
.287***
2
Adj. R
.073
.159
.167
.229
.267
2
.092
.009
.062
.039
R∆
12.843***
6.447** 47.497*** 30.923***
F∆
Note. a DV= positive behaviors; * p < .05, ** p < .01, *** p < .001

140

.542
.294***
.273
.006
5.123*

APPENDIX N
Hierarchical Regression Analysis for Research Question Three: Negative Behaviors
Table N1
Research Question Threeb: Hierarchical Regression of Demographics, Density, Factors,
Perceptions, Relations, and Sense of Belonging on Negative Behavior Participation (n=589)
Model 1
Model 2
Model 3
Model 4
Model 5
Model 6
Predictors
Step 1
Age
Sex
Race2
Race3
YrSchool
LiveOn
LiveOff
RoomLog
Step 2
AptMed
AptLow
ResHigh
ResMed
ResLow
Step 3
Factors
Step 4
Perceptions
Step 5
Relations
Step 6
SoBT

b

b

b

b

-.097
.191***
-.052
.026
-.083
.066
.137**
.271***

-.090
.194***
-.050
.043
-.084
.074
.124**
.206***

-.085
.201***
-.043
.031
-.084
.080*
.107*
.192***

-.069
.186***
-.045
.014
-.048
.047
.061
.199***

-.070
.170***
-.043
.002
-.024
.051
.043
.169***

-.070
.171***
-.043
.002
-.024
.051
.043
.169***

-.063
-.016
.198***
.042
-.060

-.065
-.010
.179***
.049
-.051

-.030
.000
.153***
.033
-.045

-.031
-.003
.118***
.018
-.048

-.031
-.003
.118***
.018
-.048

.152***

.160***

.141***

.141***

.269***

b

-.020
.424***

b

-.023
.424***
.006

R
.382
.442
.466
.533
.608
2
R
.146***
.195***
.217***
.284***
.370***
2
Adj. R
.135
.177
.198
.266
.352
2
.049
.022
.067
.086
R∆
7.022*** 16.042*** 53.810*** 77.646***
F∆
Note. b DV = negative behaviors; * p < .05, ** p < .01, *** p < .001

141

.608
.370***
.351
.000
.024

REFERENCES

142

REFERENCES

Astin, A.W. (1977). Four critical years: Effects of college on beliefs, attitudes, and knowledge.
San Francisco, CA: Jossey-Bass.
Astin, A. W. (1993). What matters in college? Four critical years revisited. San Francisco, CA:
Jossey-Bass.
Baker-Minkel, K., Moody, J. & Keiser, W. (2004). Town and gown. Economic Development
Journal, 7-15.
Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for interpersonal
attachments as a fundamental human motivation. Psychological Bulletin, 117(3), 497-529.
Blankinship, D. G. (2008, August 8). Universities try to control students off campus.
NBCNews.com. Retrieved from http://www.nbcnews.com/id/26349148/ns/us_newseducation/t/universities-try-control-students-campus/#.VBzdyytdV69
Bollen, K. A., & Hoyle, R. H. (1990). Perceived cohesion: A conceptual and empirical
examination. Social Forces, 69(2), 479-504.
Bruning, S. D., McGrew, S. & Cooper, M. (2006). Town–gown relationships: Exploring
university–community engagement from the perspective of community members. Public
Relations Review, 32, 125–130.
Chavis, D. M., & Wandersman, A. (1990). Sense of community in the urban environment: A
catalyst for participation and community development. American Journal of Community
Psychology, 18(1), 55-81.
City of East Lansing (2015a, May 1). Major Employers. Retrieved from
http://www.cityofeastlansing.com/365/Major-Employers
City of East Lansing (2015b, May 1). Quick Facts. Retrieved from
http://www.cityofeastlansing.com/367/Quick-Facts
Colorado State University (2014). Off-campus life. Retrieved from
http://www.offcampuslife.colostate.edu
Community Relations Coalition (2013). About us. Retrieved from
http://www.wealllivehere.org/about-us.html
Cossick, S. (2009, April 22). City passes noise ordinance. The Daily Athenaeum. Retrieved from
http://www.thedaonline.com/article_3d1f243b-eee4-55a0-bead-ff9bfd1771b6.html

143

Council for the Advancement of Standards in Higher Education (2012). CAS professional
standards for higher education (8th ed.). Washington, DC: Author.
DeVellis, R. F. (1991). Scale development: Theory and applications. Newbury Park, CA: Sage.
Dugan, J.P., Garland, J. L., Jacoby, B., & Gasiorski, A. (2008). Understanding commuter student
self-efficacy for leadership: A within-group analysis. NASPA Journal, 45(2), 282-310.
Florida, R. (2013, October). The boom towns and ghost towns of the new economy. The Atlantic.
Retrieved from http://www.theatlantic.com/magazine/archive/2013/10/the-boom-townsand-ghost-towns-of-the-new-economy/309460/
Freeland, R.M. (2005, May 13). Universities and cities need to rethink their relationships. The
Chronicle of Higher Education. Retrieved from http://chronicle.com
Freeman, T., Anderman, L, & Jensen, J (2007). Sense of belonging in college freshman at the
classroom and campus levels. The Journal of Experimental Education, 75(3), 203-220.
Gumprecht, B. (2003). The American college town. The Geographical Review, 93(1), 51-80.
Gumprecht, B. (2008). The American college town. Amherst, MA: University of Massachusetts
Press.
Harasta, J. (2008). Town-gown relations: University and neighborhood leaders’ perceptions of
college and community relations. (Doctoral dissertation). Retrieved from ProQuest
dissertations and theses. (UMI number: 3292899)
Hausmann, L. R. M., Shofield, J. W., & Woods, R. L. (2007). Sense of belonging as a predictor
of intentions to persist among African American and White first-year college students.
Research in Higher Education, 48(7), 803-839.
Hausmann, L. R. M., Ye, F., Schofield, J. W., & Woods, R. L. (2009). Sense of belonging and
persistence in White and African American first-year students. Research in Higher
Education, 50(7), 649-669.
Hintz, J.R. (2011). Peer educators responding to an institutional challenge: Off campus student
services. In L. B. Williams (Ed.), Emerging issues and practices in peer education (New
Directions for Higher Education, No. 133, pp. 87-95). San Francisco, CA: Wiley.
Hoffman, M., Richmond, J., Morrow, J., & Salomone, K. (2002). Investigating “sense of
belonging” in first-year college students. Journal of College Student Retention, 4(57),
227-256.
Hubbard, P. (2009). Geographies of studentification and purpose-built student accommodation:
Leading separate lives? Environment and Planning A, 41, 1903-1923.

144

Hurtado, S., & Carter, D. F. (1997). Effects of college transition and perceptions of the campus
racial climate on Latino college students’ sense of belonging. Sociology of Education, 70,
324-345.
IBM (n.d.). Pairwise vs. listwise: What are they and when should I use them? Retrieved from
http://www01.ibm.com/support/docview.wss?uid=swg21475199
International Town-Gown Association. (2013a). Mission and vision. Retrieved from
http://www.itgau.org/content.aspx?ContentID=1438
International Town-Gown Association (2013b, September). International town-gown assessment
2013. Retrieved from
http://www.itga.org/userfiles/files/ITGA%20Survey%20Assessment%20Executive%20S
ummary.pdf
International Town-Gown Association (2014, August). International town-gown assessment
2014. Retrieved from
https://www.itga.org/userfiles/files/ITGA%20Survey%20Summary%202014-3.pdf
Jacoby, B. (1989). The student as commuter: Developing a comprehensive institutional
response. ASHE-ERIC Higher Education Report, 7. Washington, DC: The George
Washington University.
Jacoby, B. (2000a). Involving commuter students in learning: Moving from rhetoric to reality. In
B. Jacoby & Associates (Eds.), Involving commuter students in learning (New Directions
for Higher Education, No. 109, pp. 81-87). San Francisco, CA: Jossey-Bass.
Jacoby, B. (2000b). Why involve commuter students in learning? In B. Jacoby & Associates
(Eds.), Involving commuter students in learning (New Directions for Higher Education,
No. 109, pp. 3-12). San Francisco, CA: Jossey-Bass.
Jacoby, B., & Garland, J. (2004). Strategies for enhancing commuter student success. Journal of
College Student Retention, 6(1), 61-79.
Johnson, D. R., Soldner, M., Leonard, J. B., Alvarez, P., Inkelas, K. K., Rowan-Kenyon, H., &
Longerbeam, S. (2007). Examining sense of belonging among first-year undergraduates
from different racial/ethnic groups. Journal of College Student Development, 48, 525-542.
Kaplowitz, S. A., & Campo, S. (2004). Drinking, alcohol policy, and attitude toward a campus
riot. Journal of College Student Development, 45(5), 501-516.
Keeling, R. P. (1999). A new definition of college emerges: Everything that happens to…a
(newly defined) student, in the context of a noisy visual ‘datascape’. NASPA Forum, 20
(5), 4–5.

145

Kember, D., Lee, K. & Li, N. (2001). Cultivating a sense of belonging in part-time students.
International Journal of Lifelong Education, 20(4), 326-341.
Kemp, R. L. (2013). Town and gown relations: A handbook of best practices. R. L. Kemp (Ed.).
Jefferson, NC: McFarland.
Kittle, D. R. (2005). Parent-child myth in university and community relations. (Doctoral
dissertation). Retrieved from ProQuest dissertations and theses. (UMI number: 3204741)
Kodama, C. M. (2002). Marginality of transfer commuter students. NASPA Journal, 39(3), 233250.
Krause, K.D. (2007). Social involvement and commuter students: The first-year student voice.
Journal of The First-Year Experience & Students in Transition, 19(1), 27-45.
Kuh, G. D. (2003). What we're learning about student engagement from NSSE: Benchmarks for
effective educational practices. Change: The Magazine of Higher Learning, 35(2), 24-32.
doi:10.1080/00091380309604090.
Kuh, G.D., Gonyea, R.M., & Palmer, M. (2001). The disengaged commuter student: Fact or
fiction? Commuter Perspectives, 27(1), 2-5.
Martin, J., & Samels, J. E. (2013). Boston, Northeastern University, and other cities and schools
are building “collegetowns”. In R. L. Kemp (Ed.), Town and gown relations: A
handbook of best practices (pp. 48-50). Jefferson, NC: McFarland. (Reprinted from
College town topics, by the International Town and Gown Association, April 2006,
Clemson, SC).
Mayfield, L. (2001). Town and gown in America: Some historical and institutional issues of the
engaged university. Education for Health, 14(2), 231-240.
Martinez, M. L., Black, M. & Starr, R. H. (2002). Factorial structure of the perceived
neighborhood scale (PNS): A test of longitudinal invariance. Journal of Community
Psychology, 30(1), 23-43.
Maestas, R., Vaquera, G. S., & Muñoz Zehr, L. (2007). Factors impacting sense of belonging at
a Hispanic-serving institution. Journal of Hispanic Higher Education, 6(3), 237-256.
McLaughlin, S. M. (2011). The effects of community building programs on student
neighborhoods adjoining the urban university campus. (Doctoral dissertation).
Retrieved from ProQuest dissertations and theses. (UMI number: 3477048)
McMillan, D. W., & Chavis, D. M. (1986). Sense of community: A definition and theory.
Journal of Community Psychology, 14, 6-23.

146

Michigan State University (1999, April 29). Alcohol Action Team: Final implementation report.
Retrieved from https://www.msu.edu/~act/finrpt5.html
Michigan State University. (2014). College life: East Lansing. Retrieved from
http://collegelifeel.msu.edu
Michigan State University (2016, May 1). MSU Facts. Retrieved from
http://msu.edu/about/thisismsu/facts.html
Michigan State University (2017). Sign Up to Live On Campus Fall 2017. Retrieved from
http://liveon.msu.edu/feature/sign-live-campus-fall-2017
Moyen, E. (2006, June 8). Town-gown relations on trial. Inside Higher Ed. Retrieved from
https://www.insidehighered.com/views/2006/06/08/moyen
Mullins Jr., R. L., & Gilderbloom, J. I. (2002). Urban revitalisation partnerships: Perceptions of
the university's role in Louisville, Kentucky. Local Environment: The International
Journal of Justice and Sustainability, 7(2), 163-176. doi:10.1080/13549830220136463
Munro, M., Turok, I., & Livingston, M. (2009). Students in cities: A preliminary analysis of their
patterns and effects. Environment and Planning A, 41, 1805-1825.
National Center for Educational Statistics (2010). Profile of undergraduate students, 2007-08.
Washington, DC: US Department of Education.
National Survey of Student Engagement. (2013). A fresh look at student engagement: Annual
results 2013. Bloomington, IN: Indiana University Center for Postsecondary Research.
Nichols, D. (1990). University-community relations: Living together effectively. Springfield, IL:
Thomas Books.
Ohio State University (2014). Neighborhood services and collaboration, Off-campus and
commuter student engagement. Retrieved from http://offcampus.osu.edu/
O’Mara, M. P. (2012). Beyond town and gown: university economic engagement and the legacy
of the urban crisis. The Journal of Technology Transfer, 37, 234-250.
Ostrove, J., & Long, S. (2009). Social class and belonging: Implications for college adjustment.
The Review of Higher Education, 30(4), 363-389.
Ott, R. L., & Longnecker, M. (2001). Statistical methods and data analysis. Stamford,
Connecticut: Thomson Learning.
Pascarella, E. T. (2006). How college affects students: Ten directions for future research. Journal
of College Student Development, 47(5), 508-520.

147

Pascarella, E., Bohr, L., Nora, A., Zusman, B., Inman, P., & Desler, M. (1993). Cognitive impact
of living on campus versus commuting to college. Journal of College Student
Development, 34(3), 216-220.
Pascarella, E. T., & Terenzini, P. T. (2005). How college affects students: A third decade of
research. San Francisco, CA: Jossey-Bass.
Porter, M. & Grogan, P. (2013). Urban renewal. In R. L. Kemp (Ed.), Town and gown relations:
A handbook of best practices (pp. 223-227). Jefferson, NC: McFarland. (Reprinted from
Leveraging colleges and universities for urban economic revitalization: An action
agenda, by the Initiative for a Competitive Inner City and CEO’s for Cities, 2002,
Boston, MA)
Remler, D. K., & Van Ryzin, G. G. (2011). Research methods in practice: Strategies for
description and causation. Thousand Oaks, CA: Sage Publications.
Rhee, B. (2008). Institutional climate and student departure: A multinomial multilevel modeling
approach. The Review of Higher Education, 31(2), 161-183.
Roe Clark, M. (2006). Succeeding in the city: Challenges and best practices on urban commuter
campuses. About Campus, 11(3), 2-8.
Silverman S. C., Aliabadi, S. & Stiles, M. R. (2008). Meeting the needs of commuter, part-time,
transfer and returning students. In S. R. Harper & S. J. Quaye (Eds.), Student engagement
in higher education: Theoretical perspectives and practical approaches for diverse
populations (pp. 223-241). Florence, KY: Routledge.
Stewart, S. S. & Rue, P. (1983). Commuter students: Definition and distribution. In S. S.
Stewart (Ed.), Commuter Students: Enhancing Their Educational Experiences (New
Directions for Student Services, No. 24, pp. 3-8). San Francisco, CA: Jossey-Bass.
Strayhorn, T. L. (2008). Fittin’ in: Do diverse interactions with peers affect sense of belonging
for black men at predominantly White institutions? NASPA Journal, 45(4), 501-527.
Strayhorn, T. L. (2012). College students’ sense of belonging: A key to educational success for
all students. New York, NY: Routledge.
Sungu-Eryilmaz, Y. (2013). Evolving relations. In R. L. Kemp (Ed.), Town and gown relations:
A handbook of best practices (pp. 7-9). Jefferson, NC: McFarland. (Reprinted from
Town-gown collaboration in land use: Policy focus report by the Lincoln Institute of
Land Policy, 2009, Cambridge, MA)
Tabachnick, B. G., & Fidell, L. S. (2007). Using multivariate statistics. Boston, MA: Pearson
Education.

148

Thelin, J. R. (2004). A history of American higher education. Baltimore, MD: The Johns
Hopkins University Press.
Tinto, V. (1993). Leaving college: Rethinking the causes and cures of student attrition (2nd ed).
Chicago, IL: University of Chicago Press.
Twohey, M. (2007, March 27). Students off-campus behavior to be targeted. Journal Sentinel.
Retrieved at http://www.jsonline.com/news/milwaukee/29470154.html
University of Colorado Boulder (2011). Off-campus housing & neighborhood relations.
Retrieved from http://offcampushousing.colorado.edu/
University of Florida (2013). Off-campus life. Retrieved from http://www.offcampus.ufl.edu
Warfield, W. (1995). Town and gown: Forums for conflicts and consensus between universities
and communities. New Directions for Higher Education, 92, 63-69.
Whitson, B. (2013). Williamsburg and the College of William and Mary form joint
neighborhood relations committee. In R. L. Kemp (Ed.), Town and gown relations: A
handbook of best practices (pp. 219-222). Jefferson, NC: McFarland. (Reprinted from
William & Mary News, by the College of William and Mary, July 6, 2011, Williamsburg,
VA)
Young, C. E. (2013). Gainesville and the University of Florida implement neighborhood action
plan. In R. L. Kemp (Ed.), Town and gown relations: A handbook of best practices (pp.
84-93). Jefferson, NC: McFarland. (Reprinted from College town topics, by the
International Town and Gown Association, September 18, 2002, Clemson, SC)

149