1 THE EFFECT OF PART CERTIFICATE COMPLETION IN TWO YEAR COMMUNITY COLLEGES By HONGWEI YU A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLME NT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2013
2 2013 Hongwei Yu
3 To my wife, family members, and friends
4 ACKNOWLEDGMENTS I would like to take this opportunity to ex press my sincere appreciation to my dissertation committee members, Dr. Dale Campbell, Dr. Bernard Oliver, Dr. David Miller, and Dr. Amir Erez. Their insightful suggestions and recommendations are integral to the completion of my dissertation. I am especia lly indebted to my academic advisor and committee chair, Dr. Dale Campbell who serves as a professional mentor and caring friend during my academic studies in the Higher Education Administration program at the University of Florida Spec ial thanks to Dr. P ilar Mendoza, Dr. Dale Campbell, and Institute o f Education Sciences, who allow me to access to the restricted BPS dataset. Special t hanks to Dr. David Honeyman for his encouragement during my academic studies. My gratitude and appreciation also go to my wife, Wei Xu, whose understanding, support, and sacrifices make my dream become a reality. The dissertation would be impossible if you are not around. Thanks for the support and encouragement from my parents and family members. To Almighty God who provide s love, blessing, strength, and confidence along the road of completing this dissertation and everything else
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 8 ABSTRACT ................................ ................................ ................................ ..................... 9 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 11 Benefits of Hi gher Education ................................ ................................ .................. 11 The Critical Role of Community College ................................ ................................ 12 Student Success in Community College ................................ ................................ 14 Statement of Problem ................................ ................................ ............................. 15 Research Questions ................................ ................................ ............................... 18 Definitions of Terms ................................ ................................ ................................ 18 Significance of the Study ................................ ................................ ........................ 19 Chapter Summary ................................ ................................ ................................ ... 20 2 THE REVIEW OF LITERATURE ................................ ................................ ............ 21 Community College Context ................................ ................................ ................... 21 Community College History ................................ ................................ .................... 23 The Expansion of Secondary School Era (1900s 1940s) ................................ 24 The Great Expansion of Community Colleges (1950s 1970s) .......................... 24 Comprehensive Community Coll ege (1980s 1990s) ................................ ........ 25 The Advanced Community College System (1990s Present) ........................... 26 Part time Faculty in Community Colleges ................................ ............................... 26 Part time Faculty in Higher Education Institutions ................................ ............ 26 Community College Faculty ................................ ................................ .............. 27 Part time Faculty in Community Colleges ................................ ......................... 27 Part .......... 28 Student Degree or Certificate Completion ................................ .............................. 34 Research on Student Degree or Certificate Completion ................................ ... 34 Theoretical Models Predicti ng Student Dropout or Retention ........................... 35 Critical Review on Theoretical Models on Student Retention ........................... 38 Variables Impacting Student Deg ree or Certificate Completion .............................. 39 Student Demographics and Prior Experiences ................................ ................. 40 Student College Experiences ................................ ................................ ........... 42 Institutional Characteristics ................................ ................................ ............... 50 Chapter Summary ................................ ................................ ................................ ... 53
6 3 METHODOLOGY ................................ ................................ ................................ ... 54 Research Design, Datasets, and Sample Size ................................ ....................... 55 Data Source and Variables ................................ ................................ ..................... 57 Mu ltiple Imputation ................................ ................................ ................................ .. 59 Logistic Regression ................................ ................................ ................................ 61 Multi level Modeling: Concepts and Application ................................ ...................... 61 Centering Option ................................ ................................ ................................ ..... 64 Weighting Issue ................................ ................................ ................................ ...... 65 Limitations of the Study ................................ ................................ ........................... 65 Chapter Summary ................................ ................................ ................................ ... 66 4 RESULTS ................................ ................................ ................................ ............... 67 Statistical Software: R, HLM, SPSS ................................ ................................ ........ 68 Descriptive Statistics ................................ ................................ ............................... 69 Unconditional Analysis ( Random ANOVA ) ................................ ............................. 70 Chapter Summary ................................ ................................ ................................ ... 76 5 CONCLUSIONS AND IMPLICATIONS ................................ ................................ ... 79 Institutional Characteristics ................................ ................................ ..................... 79 Individual Level Variables ................................ ................................ ....................... 81 Student Demographics and Prior Experiences ................................ ................. 81 Student College Experiences ................................ ................................ ........... 83 Data ................................ ................................ ................................ ..................... 84 Implications ................................ ................................ ................................ ............. 86 Implications for Higher Education Research ................................ ..................... 86 Implications for Higher Education Administration ................................ ............. 87 Concluding Though ts ................................ ................................ .............................. 88 APPENDIX A Criteria for Selecting Community Colleges ................................ ............................. 90 B R Code for Combing Datasets and Normalizing Weights ................................ ....... 92 REFERENCE LIST ................................ ................................ ................................ ........ 96 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 114
7 LIST OF TABLES Table page 4 1 Dependent and independent variables used in the analysis ............................... 67 4 2 Descriptive statistics for level 1(student level) variables ................................ ..... 69 4 3 Descriptive statistics for the level 2(institutional level) variables ......................... 70 4 4 Final estimation of variance components for pratt3yij(unconditional) ................. 71 4 5 Final estimation of variance components for pratt6yij(unconditional) ................. 71 4 6 Final estimation of fixed effects ................................ ................................ .......... 72 4 7 Final estimation of fixed effects ................................ ................................ .......... 75 4 8 Final estimation of variance components for pratt3yij (conditional) .................... 77 4 9 Final estimation of variance components for pratt6yij (conditional) .................... 78 5 1 ....................... 86
8 LIST OF FIGURES Figure page 2 1 The conceptual framework predicting student degree or certificate completion in two year community colleges ................................ ................................ .......... 39
9 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy THE EFFECT OF PART TIME FACULTY ON STUD CERTIFICATE COMPLETION IN TWO YEAR COMMUNITY COLLEGES By Hongwei Yu August 2013 Chair: Dale Campbell Major: Higher Education Administration Featured as open acce ss, low cost tuition and fees, geographical proximity, community college has become the largest sector in American higher education with diverse student body and multiple missions (Cohen & Brawer, 2008 ; College Board, 2010 ). One of the most significant trends in community colleges involves the presence of part time faculty on comm unity college campuses (Jacoby, 200 6). Prior research studies linked the employment of part time faculty with student degree or certificate completion (Benjamin, 2002; Ehrenberg & Zhang, 2005; Jacoby, 2006; Leslie & Gappa, 2002 ; Umbach, 2008; Umbach & Wawr zynski, 2005). However, the research literature on this topic yields inconsistent research findings. In this study, I draw the datasets from the Integrated Postsecondary Education Data System (IPEDS) and The Beginning Postsecondary Students Longitudinal S tudy (BPS:04/09), and employ a multi level logistic modeling approach to present a comprehensive picture about the effect of part time faculty on individual student probability of degree or certificate completion in two year community colleges. The analyti cal results indicate that part time faculty have a negative impact on
10 completion. Students who study in community colleges located in rural areas are less likely to graduate than their peers in community colleges located in non rural places Institution size is inversely related to student degree or certificate completion The current study also confirms that student academic preparation contributes to student degree or certificate completion Although the majority of community college s ha ve no admission requirement s about student academic performance, having better academic preparation (i.e. high school GPA ) appears to have a significant and positive impact on likelihood of degree or certificate completion Full time attendance in community colleges is also positively associated with student degree or certificate attainment. The study also shows that students who spend much time working instead of study ing are less likely to finish the ir degree or certificate. In addition, college tuition and fees ha ve a small positive impa The implications for higher education research a nd administration are also presented.
11 CHAPTER 1 INTRODUCTION How do you des cribe your feelings when you explore the adventures and parades of the Disneyland but have to leave during the process? You may think going to the Disneyland is not a wise decision as you waste your time and money. This scenario is similar to attending hig her education without completing degrees or certificates. Students who dropout from institutions of higher education may not only miss the opportunity to explore the college life but miss the chance to make a good living during their life. Such loss is cer chance to enjoy oneself in the Disneyland. I dentifying factors tha t hinder students from completing their degree or certificate is of great significance. This is particularly true for two year c ommunity co lleges where a significant proportion of students leave these two year institutions of higher education without completing associate degrees or certificates ( Provasnik & Planty, 2008 ) Benefits of Higher Education One thing higher education researchers ra rely disagree is that pursuing higher education is a worthwhile investment for both individuals and society as a whole (College Board, 2010; Pascarella & Terenzini, 1999, 2005; Perna, 2005 ). Indeed, higher 2005). Receiving higher education is associated with how individuals make a living. Every three years, the College Board publish es its widely cited report Education Pays reporting the differences in earning s health benefits, and societal benefits between students who attend higher education and students who only receive high school diploma per
12 year 21,900 dol lars more than students who only gradu ate from high school ( College Board, 2010) The average expected lifetime earnings for a student with an associate degree are 1.6 million dollars, a 0.4 million more than a high school graduate (Mellow & Heelan, 2008). In addition, individuals who attend higher education are more likely to receive health benefits and are more satisfied with their work than their high school counterparts Society benefits by collecting increased tax revenues from college graduates and spending less in sup port programs for t hese individuals (College Board, 2010). In addition to these tangible economic benefits, attending higher education is Pichler 2005): individuals who receive higher edu cation are more likely to lead a healthier lifestyle as compared to their higher school counterparts. In addition, higher education can also bring the so individuals are more likely to help their children p repare for the future academic studies than their high school counterparts. Furthermore, college educated individuals are more likely to be involved with volunteering activities and community services (College Board, 2010; Paulsen, 1996). As such, we have witnessed an enrollment surge in institutions of higher education. According to the Digest of Education Statistics 2011 college student enrollment in degree granting institutions has increased 11% between the year 1990 and the year 2000 ; whereas the col lege enrollment between the year 2000 and the year 2010 have increased 37%, from 15.3 to 21.0 million ( Snyder & Dillow, 2012 ). The Critical Role of Community College Although attending higher education has significant imp act on individuals and society a s a whole, many students are excluded from the door of four year colleges or
13 universities. As researchers pointed out, there is a substantial gap in terms of higher education enrollment between various racial/ethnic groups as well as socio economic groups (Bowen, Chingos, & McPerson, 2009). Community college s play an indispensable role in matriculating four year institutions. Community college s enroll more than half of undergraduates in public higher educati on ( American Association of Community Colleges (AACC), 2013; Schuetz, 2005). These institut ions of higher education are characterized as ope n access, convenient location and low tuition. And they allow students to attend postsecondary education and offer various career paths for diverse students (AACC, 2013) Within community colleges, t here is an over representation of minority students, low income students, and first generation students. Upon examining fast growing jobs in four sectors of American econo my, Milano, Reed, and Weinstein (2009) reported jobs that requiring two year associate degrees would grow nearly twice as much as the national average and the demand for community college graduates would be larger than that for graduates from four year ins titutions For many students community colleges are the dream come true places. Research studies also highlighted the role of community colleges in American higher education by mentioning the American Graduation Initiative l aughed by President Barack Obam a, which call ed on community c olleges to have additional five million graduates by the year 2020 ( Alexander, Karvonen, Ulrich, Davis, & Wade, 2012; Rose & Hill, 2013). Such initiative will help United States regain its leaders hip position in higher educati on completion rate s (Rose & Hill, 2013).
14 Student Success in Community College As indicated above, student degree or certificate completion has a great impact upon individuals and society. While student enrollment in institutions of higher education increa sed dramatically since the mid 1900s to about 14 million students each year, a considerable nu mber of students who entered achieve degree or certificate completion (Swail, 2004). Despite the rising enrollment in in st itutions of higher education student graduation rate in four yea r institutions changed little. According to American College Testing (ACT) 2012 report, nearly half of undergraduate students in four year degree completion ever since the year 1991(ACT, 2012). The low graduation rate also rings true in two year community colleges. Of all first time students who were enrolled in community colleges in 1993, only 36% complete their certificate s associate or bachelor degrees within six years (Bailey & Alfonso, 2005). About 45% of students who started community college s in 2003/ 04 left their school s without completing an associa te degree or certificate by the year 2006 (Provasnik & Planty, 2008). More recent report indicated that nea rly half of students leave the community colleges without completing an associate degree or a certificate or transf erring to four year universities within six years (U.S. Department of Education, 2011) ave a n academic goal of degree or certificate completion, such low degree or certificate completion rate is nevertheless considered as unsatisfying (Bailey & Leinbach, 2005 as cited in Bailey & Alfonso, 2005) How to ensure student degree or certificate co mpletion beyond student access to post secondary education is of vital importance for higher education
15 administrators and researchers who are determined to help students reap the tangi ble and intangible benefits of higher education. Statement of Problem Re searchers maintained that individual background and institutional environment probability of academic attainment (i.e. Bailey, Calcagno, Jenkins, Kienzl, & Leinbach, 2005 ; Jacoby, 2006 ; Pascarel la & Terenzini, 1991, 2005 ). An important line of research involves the investigation of part time faculty on student degree or certificate completion in institutions of higher education. Faculty play an integral role in ensuring student degree or certificate completion in institu tions of higher education (Astin, 1999 ; Pascarella & Terenzini, 1991). Through their instructional activities, faculty influence the curriculum and help foster the environment for student learning ( Boyer, 1987). Over the past thirty years, the increasing p resence of part time faculty has become one of the most salient trends in institutions of higher education in general and i n community colleges in particular (Jacoby, 2006). In the year 2003, part time faculty account for 44% of faculty and instructional staff in institutions of higher education ( Cataldi, Fahimi, Bradburn & Zimbler, 2005) and teach 37% of college undergraduates (Twombly & Townsend, 2008) Currently, part time faculty account for two thirds of faculty and teach one third of all course secti ons in two year community colleges ( Twombly & Townsend, 2008 ). Part time faculty is more represented in community colleges than in four year institutions ( Curtis & Jacobe 2006). In fact, part time faculty account for 6 7 % of faculty in commun ity colleges a s compared to 22% to 25% of faculty in other types of higher education institutions ( Cataldi, Fahimi, Bradburn & Zimbler, 2005 ).
16 Although student attrition has been a longstanding challenge for community colleges, there has been little research examining the effectiveness of full time faculty or instructors on student degree or certificate completion Yet, there has been extensive research on the effect of part time faculty degree or certificate completion ( Benjamin, 2002; Bettinger & Long, 2006, 2010; Calcagno, Bailey, Jenkins, Kienzl, & Leinbach, 2008; Eagan & Jaeger, 2008; Ehrenberg & Zhang, 2005; Jacoby, 2006; Jaeger & Eagan, 2009; Titus, 2004; Umbach, 2007, 2008). An important reason for studying part time on student deg ree or certificate completion is the fact that part time faculty serve as an inexpensive alternative for full time faculty members. They generally receive less compensation and other benefits, and are often denied access to offices and computers. Researche rs are concerned whether they can perform as effective as their full time counterparts given the fact that part time faculty are less integrated into these two year institution s (Tinto, 2006). Indeed, examining the relationship between part time faculty and student probability of degree or certificate completion in community colleges become the central focus of current literature (Benjamin, 2002; Bettinger & Long, 2010; Calcagno, Bailey, Jenkins, Kienzl, & Leinbah, 2008; Cohen & Brawer, 2003; Eagan & Jae ger, 2008; Ehrenberg & Zhang, 2005; Jacoby, 2006; Jaeger & Eagan, 2009; Sandy, Gonzalez, & Hilmer, 2005; Umbach, 2007). Researchers attributed the unsatisfying degree or certificate completion rate to the increased employment of part time faculty. Critics such as Benjamin (2002) maintained that student learning might be compromised by being expos ed to part time faculty in that part time faculty are less engaged with their students and/ or use less challenging instruction methods than their
17 full time counter parts (Benjamin, 2002). However, the data and analysis Benjamin presented is not sufficient to draw the convincing conclusion that part time faculty has a nega tive impact on student learning outcomes (Benjamin, 2002). Jacoby (2006) studied th e relationshi p between employment of part time faculty and student degree completion in community colleges. Through analyzing the data from The Integrated Postsecondary Education Data System (IPEDS), Jacoby (2006) found employing a higher percentage of part tim e facult y has a significant negative impact on student associate degree completion. However, he constructed the model by aggregating both faculty and student level data to institutional level since the IPEDS dataset only offers institutional level data of two year community colleges. As such, individual d ifference is inevitably masked. Such analysis might result in inaccurate parameter estimates and standard errors (Raudenbush & Bryk, 2002). Jaeger and Eagan (2009) built a multi level model by including both inst itutional and student leve l variables to investigate whether part time faculty have an impact on student degree completion in California community college system T heir study separated individual level variances from institutional level variance s Their re search results indicated reduced likelihood to graduate is more likely to be due time faculty members rather than due to the o verall percentage of part time faculty employed by a community college. Ho wever, their analytical results are based on data collected in California community colleges. include
18 Research Quest ions T hrough using IPEDS and BPS data sets, the current research intends to offer a comprehensive picture about part time f or certificate completion in two year community colleges. Multi level logistic model ing is employed to better account for non independence of variances across different community colleges. Specifically, the researcher intends to answer the following research questions: 1. Whether part time faculty significantly impact student degree or cer tificate completion while holding individual characteristics and other institutional level variables constant. If there is a significant effect on student degree or certificate completion, the magnitude and direction of such effect. 2. Whether individual lev el variables and other institutional variables significantly impact student degree or certificate completion. If there are significant effects, the magnitude and the direction of these effects 3. Are there any difference s in terms of variables that impact s tudent degree or certificate completion Definitions of Terms 1. Community colleges : In line with the IPEDS dataset, community colleges are defined as community colleges offering associate's degree and certificate programs. They also include two year institutions that award baccalaureate degrees where the percentage of baccalaureate degrees is less than ten percent of all degrees awarded (IPEDS, 2013) 2. Student academic achievement : While some researchers believe s tudent academic achievement education goals (AFT, 2011), the current study defines student academic achievement f completing associate degree s or certificate s in two year community colleges with in three or six years. 3. Part time faculty: In the current study, p art time faculty refer to faculty serving in a temporary capacity to teach courses. Part time faculty also include faculty who are hired to teach remedial, developmental, or ES L course s (IPEDS, 2013)
19 Significance of the Study The current study adds new understand ing about the impact of part tim e faculty on student degree or certificate completion in two year community colleges. Built upon previous research studies, various ind ividual and institutional level characteristics are examined to see whether these variables have significant impact on student degree or certificate completion in two year community colleges. In particular, the current study investigates whether employing part time faculty has a significant negative impact on student degree or certificate completion as indicated by prior studies. The combination of the public released IPEDS dataset and the restricted version of BPS: 04/09 dataset coupled with the usage of multilevel logistic modeling enab le the researcher to provide additional findings on this research topic. The current research also provides some implications for the educational practitioners and policy makers. Each year a considerable amount of resourc es ha ve been invested in community colleges to retain students. However, there are still a high percentage of students drop ping out of community colleges ( Aud et al., 2012 ). To make the community college a dream come true place, policy makers and college a dministrators should work collaboratively to make sure students achieve degree or certificate completion. The current research identifies a group of individual and institutional level variables that explain why these students leave their institutions whic h can assist practitioners to make informed decisions regarding how to invest effectively to make sure students have achieve degree or certificate completion Educational
20 C hapter Summary Student degree or certificate completion plays a central role in ensuring students to reap the benefits of higher education and student degree or certificate completion in community college s is of great significance where almost half of und ergraduate students start their post secondary education With hi gh enrollment and low degree or certificate attainment community college s still have to do more work to help students achieve their dream. Based on theoretical and empirical exploration s of previous studies the current study investigates the effect of individual and institutional level variables on student academic achievement with a primary focus on the impact of overall percentage of part time faculty on student degree or certificate compl etion The study draws data from two national datasets developed by National Center for Education Statistics: IPEDS and BPS: 04/09, which adds new understanding about the effect of hiring part time facult y and other variables that play an essential role fo r student degree or certificate completion Multilevel modeling is used to account for variances across different community colleges. A review of literature is followed to critically examine the current research on community colleges, community college fac ulty (part time faculty in particular) student degree or certificate completion and student and institutional l evel variables that were found to have an impact upon student degree or certificate completion
21 CHAPTER 2 THE REVIEW OF LITE RATURE To inform this current study, I review the relevant research on student degree or certificate completion in institutions of higher education. I start the literature review by discussing t he critical role community college s play in American higher ed ucation and their diverse student population and multiple education missions Then, I provide a succinct chorological review of community college s: how they evolve and exp and ever since their inception. Thirdly I provide a brief introduction about the f a culty members in com munity college s with a primary e mphasis on part time faculty and their impact on student degree or certificate completion Then, I review theoretical models pertain to student persistence that were developed in institutions of higher e ducation and propose a retention framework for community college students Finally, I review student and institution level variables that are included in proposed student retention framework Community College Context Community college s are a unique inven tion in American higher education (Beach, 2011). Originally, they have a strong emphasi s on student academic transfer (Dougherty, 1994 as cited in Eagan & Jaeger, 2008), helping students prepare for their future studies in four year institutions. However, their mission gradually expands to encompass developmental and remedial education, continuing education, community services, and voc ational and technical training (Cohen & Brawer, 2008). F ew of their conceivers would have imagined that community college s h a ve such an expansion over a century. Community college s ha ve become the largest secto r of American higher education system (Boggs, 2012). C ommunity college s offer many students a precious opportunity to attend post secondary education (Cohen & Brawer,
22 200 3, 2008 ; Pascarella 1999 ). And many students take the advantage of relatively low cost in two year community colleges to finish the first two years of coursework toward a year institutions (American Associatio n of Community Colleges (AACC) 2013). Students can also find programs that provide knowledge and skills in professional areas. For instance, community colleges educate nearly 60% of the new nurses, 80% of firefighters, law enforcement officers, and emerge ncy medical technicians (EMTs) ( AACC 2013). Community college s have a diverse and large student body. Community college s enroll a high percentage of non traditional students ( Cohen & Brawer, 2003; Jaeger & Eagan, 2009). For instance, t hey enroll 58% of pa rt time students, 42% of first generation students, 13% of single parents, 6% of non U.S citizens, 3% of veterans, 12% of students with disabilities, 43% of the first time freshmen, 51% of Hispanic students, 44% of African American students, 54% of Native American students, and 45% of Asia n/Pacific Islander students ( AACC 2012 ; Planty et al 2007 ). Currently, almost half of undergraduate students( there are more than 11.5 million community college students, with 6.5 million students enrolled in credit cou rse work s ) start post secondary education in 1,132 community colleges, including 986 public community colleges, 115 independent colleges, and 31 tribal colleges ( AACC 2012). Several reasons lead to the enrollment surge in community colleges: a higher perc entage of high school graduates, economic and social benefits of degree or certificate completion, increasing participation of non traditional students, fem ale students, and minority students in post secondary education (Hagedorn, 2010).
23 In sum, community college s play an essential role in pro moting student access to post secondary education, without which many would have had little opportunity to attend higher education (Townsend & Twombly, 2001). Yet, degree completion for community college students is g enerally unsatisfying ( Calcagno, Bailey, Jenkins, Kienzl, & Leinbah, 2008 ; Rosenbaum, Deil Amen, & Person 2006). E xpanding access without p students fully reap the benefits of higher education How to ensu re student academic achievement becomes one of the most urge nt issues in community colleges. Community College History The appearance of two year junior colleges is directly associated with four year first two years in four year institutions In the eyes of early higher education administrators and researchers community college s allow high school student four year institutions can focus on research (Cohen & Brawer, 2008 ). Educational leaders from four year institutions such as Henry Tappan, the then president of University of Michigan, William Folwell, the then president of University of Minnesota maintained that four year institutions should concentrate on highe r level scholarship, and stay away from the preparatory coursework. They conceived new educational insti tutions be established to provide freshman and sophomore years of course ; whereas four year institutions could focus on research activitie s (Cohen & Brawer, 2008). T he deeper reason for the inception of community college s is societal response to educational needs and social pressure for housing an increasing number of high school graduates who were looking for opportunities to receive higher educatio n (Beach, 2011; Cohen, 1998).
24 Joliet Junior College was the first community college established in American higher education. The college was cofounded in 1901 by William Harper, the then president of University of Chicago and Stanley Brown, the then prin cipal of Juliet High School (Hagedorn, 2010). Ever since their inception community college s ha ve experienced several stages of expansion. And community colleges experience great expansion during 1920s, late 1930 and 1 940s, and 1960s (Beach, 2011). The pri mary reason for such expansion is the increasing number of students looking for post secondary education opportunities. The Expansion of Secondary School Era (1900s 1940s) From their inception to 1940s, community college s are commonly known as junior coll ege s The majority of junior colleges were private colleges. In the first decade of the 20 th century, there were only 25 junior colleges and the number increases to 325 colleges in the 1927 with a total enrollment of 35,630 students (Beach, 2011). Most of these institutions were really small with a total enrollment of 20,000 students. By 1930, there were 450 junior colleges with a total enrollment of 70,000 students. The average institution size was still small about 160 students per institution (Cohen & B rawer, 2008). California was the leading state in establishing public junior colleges with many colleges established in Illinois, Texas, and Missouri. Between 1930s and 1940s, the junior college s started offering courses in occupational areas as a way to e liminate high unemployment rate ( AACC 2012). At the end of this era, there were 575 colleges in all 45 states with an average institution size of a couple of hundred students (Geller, 2001). The Great Expansion of Community Colleges (1950s 1970s) The Se and Higher Education for American Demo cracy 1946 promoted the expansion s of
25 community college s These policies recognized the importance of two year education in community colleges. The latter one called for the establishment of community colleges so that young people can benefit. However, critics believed these policies were designed to keep a considerable number of labors in the educational pipeline to prevent unexpectedly high unemployment r ate (Brint & Karabel, 1989). In the 1960s, community colleges experienced a swift growth with some 50 new colleges established each year (Cohen, 2001) This is directly related to state involvement in expanding community colleges and California, Florida, I llinois, Michigan, and North Carolina became the states with the most comprehensive community college system (Cohen, 2001). In 1971, 847 public junior colleges enrolled more than 2.1 million students with 244 private junior colleges enrolled 131,000 studen ts. Together, they enrolled about a third of all undergraduate students (Beach, 2011). The number of public colleges tripled in the twenty years prior to 1975, growing from 336 to 981 (Cohen, 1998). The public community colleges showed the most significant growth with an enrollment of five million students in 1975. Comprehensive Community College (1980s 1990s) During this time, community college s experienced modest growth and evolved into a mature post secondary education system. Community college s enrolle d 45% of all first time students and more than 20% of these students transfer to four year institutions (Cohen, 1998). As such, four year colleges or universities relieved the burden of hosting an increasing number of high school students and were able to maintain selective admission standards (Cohen, 1998). Community colleges have transformed from college preparatory institutions into comprehensive colleges addressing diverse needs of local communities. The late 1980s and 1990s witnessed
26 the new trend of w orkforce development and customized training, which led to greater mission expansion for community colleges across the nation. The Advanced Community College System (1990s Present) Community colleges have greatly expanded in terms of student enrollment. A s of 2011, there are 44% of undergraduate students enrolled in community colleges and about eight million students take credits from these postsecondary education institutions (AACC, 2012). Community college s have shifted their focus from student learning and meeting local needs to workforce development tha t is adaptive to global economy (Levin, Kater, & Wagoner, 2006).Community college s have not only survived but thrived by demonstrating their flexibility an d effectiveness in addressing diverse community n eeds (AACC,2012). Part time Faculty in Community Colleges Part time Faculty in Higher Education Institutions Over the past three decades, there has been a significant increase in the number of part time faculty teaching in the institutions of higher educ ation. While part time faculty accounted only one third of the faculty in the 1975(Monks, 2009 ), they nevertheless account for half of faculty population in the year 2009 ( Aud et al 2011). T here were a total of 1.4 million faculty which includes 0.7 mill ion part time faculty and 0.7 million full time faculty ( Planty et al 2009). One of the most noticeable differences between part time and full time faculty are their academic credentials (Cohen & Brawer, 2008; Levin 2012). While two thirds of full time faculty hold a Ph. D or a professional degree in four year colleges or universities, only 27% of part time faculty hold such degrees in two year community colleges (Monks, 2009). The majority of part time faculty teach at one institution: 79% of
27 part time faculty indicate they only have one teaching job, while 17% of part time faculty teach at one other institution and only 4% of the part time faculty teach at two or more other jobs (Monks, 2009). Community College Faculty Faculty members are more gender balanced in community colleges than faculty members in other institutions of higher education (Twombly & Townsend, 2008). It is estimated that 47% of faculty teach liberal arts, 40% teach in professional fields, 8% teach in vocational areas, and 4% teach i n developmental education (Levin, Kater, & Wagoner, 2006). Minority faculty are underrepresented given they account for only 20% of faculty working in two year community colleges (Levin, Kater, & Wagoner, 2006). The primary work for community college facul ty is teaching. Almost 90% of their time is devoted to teaching as compared to 63% of time of their counterparts at four year institutions is devoted to teaching. And community colleges faculty have heavier course loads and teach more students than their four year counterparts (Townsend & Twomby, 2007). Unlike most four year institutions community colleges provide little support for faculty to conduct research (Twombly & Townsend, 2008). Part time Faculty in Community Colleges Part time faculty are a sig nificant component of community college faculty, their number increased dramatically between 1980s and 1990s. This trend is consistent with the increasing reliance on part time workers in the new economy ( Levin, Kater, & Wagoner, 2006). Part time faculty a ccount for 70% of community college faculty (AFT, 2011 ) and teach about one third of the courses in community colleges. In addition, community college faculty members teach around 37% of college students, which include 50% of freshman and sophomores (Twomb ly &Townsend, 2008).
28 An important reason why community colleges rely on part time faculty is that hiring part time faculty is much cheaper than hiring a full time faculty (Levin, Kater, & Wagoner, 2006; Umbach, 2007; Cohen & Brawer, 2008). In fact, hiring a part time faculty can be 80% cheaper than hiring a full time faculty (College and University Professional Association for Human Resources, 2001). The sharp decrease in state funding and the dramatic increase in student enrollment all the more strengthen part (Cohen & Brawer, 2008 ). In addition, part time faculty can be employed or dismissed according to institutional needs. Indeed, their relationship with community colleges is similar to that of migran t workers with farms (Cohen & Brawer, 2008). Prior to 1970s, part time faculty were mainly qualified instructors in high school or well known profess ors in four year institutions. By contrast, the current part time faculty tired teachers, business or professional people) or presence of part time faculty is hypothesized to bring changes in decision making process as fewer faculty will be involv ed in the process Furthermore, the over representation of part time faculty is believed to have an impact on student probability of degree or certificate completion in two year community colleges ( Benjamin, 2002; Cohen & Brawer, 2008). Part s Impact on Student Degree or certificate completion Part class Center for Education Statistics indicate d part time faculty are the majority of faculty in community colleges (Leslie & Gappa 2002) Part time faculty are hired for different
29 purposes: liberal arts faculty are hired primarily for their labors as substitutes for full time faculty; while occupati onal related faculty are employed for their specialized knowledge(Levin, 2007 as cited in Jaeger & Eagan, 2009). Their increasing employment in institutions of higher education has either been hailed as an effective administrative strategy to cope with cha nging environment or criticized as an unfeasible way to achieve desirable educational outcomes (Calcagno, Bailey, Jenkins, Kienzl, & Leinbah, 2008; Cohen & Brawer, 2008; Eagan & Jaeger, 2008; Gappa & Leslie, 1997; Harrington & Schibik, 2001; Jaeger & Eagan 2009; Johnson, 2006; McLaughlin, 2005). Positive impact on student learning outcomes Their increasing presence has been embraced and welcomed by advocate s such as Gappa and Leslie (1993, 1997 ). They maintained part time faculty were committed professio nals motivated by the ir personal interest in instructional practices Part time faculty, especially those in technical programs, can bring fresh working experience and expertise to classroom, which full time faculty may not possess. Programs such as workfo rce training, adult education, and English as a second language benefit from p art experience s and expertise (Armstrong, 2005; Charlier & Williams, 2011; McLaughlin, 2005). Instruction given by these part time faculty allows communi ty colleges to keep in pace with industrial needs. These part time faculty might be employed in the industry, which is helpful for their students to have internships or even jobs, In addition, using part time faculty enables full time, tenured faculty to c oncentrate on their research in the four year colleges or universities (Bettinger & Long, 2010). Employing part time faculty is an effective way to meet the enrollment demands under the budgetary cuts (Charlier
30 & Williams, 2011; Ehrenberg & Zhang, 2005) an Bettinger and Long (201 0) investigated whether part time faculty have an impact data set that contai ns application and transcript information for over 43,000 students in Ohio public four year universities between fall 1998 and fall 1999. Their finding revealed that part time faculty ha d a posi tive impact on student enrollment in particular fields. Part t ime faculty tend to have either positive or no effect upon student academic interest. In particular, p art time faculty in occupation related fields ( i.e. educ ation, engineering, and sciences ) have s and ma jor choices. Johnson (2006) observed the effect of exposure to part time faculty tends to be minimized when taking student characteristics into the account and point ed out that the high presence of part time faculty in community colleges was related to stu when analyzing part act on student degree attainment in community colleges. Negative impact on student learning outcomes On the other hand, part time fac hiring part time faculty was no longer the institutional practices that bring new knowledge and strength en capacity; rather such practice is more and more implem ented as a simple alternative to weather the budgetary constraints (Jacoby, 2006). And hiri ng a high percentage of part tim e faculty is negatively associated with
31 student degree or certificate completion in higher educatio n (Benjamin, 2002; Egan, 2008 ; Ega n & Jaeger, 2008; Ehrenberg & Zhang, 2005; Jaeger & Egan, 2009). Researchers such as Benjamin (2002) questioned the instructional quality of part time faculty. His research indicated that people with advanced degrees such as Ph.D. degrees used different hold such advance degrees Part time faculty usually do not have such advanced degree s and in their instructional activities ; they are les s likely to challenge students academically in the class ; and they spend relatively les s time preparing for their instruction (Umbach, 2007). As such the over reliance on part time faculty shaped the undergraduate experience and was detrimental to student learning. Another reason related to the negative impact of part time faculty is grading differences between part time and full time faculty. Part time faculty are more likely to inflate their grading when part cont ract is not long term and administrators have little information about part time faculty other than student evaluation to decide whether they should rehire these part time faculty(Jacoby, 2006). This is because low student grading is often inversely associ ated with good student evaluations (Greenwald & Gilmore, 1997 as cited in Jacoby, 2006). As such, it is reasonable that part time faculty are more likely to provide less rigorous course s and provide student with inflated grades in exchange for favorable s tudent evaluations. unsupportive organizational climate and low job satisfaction (Weiss & Pankin, 2011). Compared to their full time counterparts, part time faculty members are less
32 compensated. While full time faculty report the average earning for instruction is $46,636 dollars, average earning from instruction for part time faculty is only $9,782 dollars. Earning from instruction for full time faculty is nearly five times as m uch as that for part time faculty. In addition to low rates of compensation, they are less likely to be eligible to participate in college benefit plans (Jacoby, 2006). Their relatively low wage and benefits often leave part time faculty no choice but spe nding more time teaching at multiple institutions. As such, they have little time engaging with students in one single institution (Benjamin, 2002). Part time faculty often have limited access to offices and computer s (Weiss & Pankin, 2011), and are unlike ly to obtain support for professional development (Benjamin, 2002). In a word, community college part time faculty are less likely to have a sense of organizational attachment due to the unsatisfying working conditions (Eagan & Jaeger, 2008). Harrington a nd Schibik (2001) studied the relationship between part time faculty usage and student retention in a comprehensive university. The researchers matched instructor, residen tial status.) with instructor types (part time faculty vs. full time faculty). time faculty in their first semester and their retention rate in their second semester. Ehrenber g and Zhang (2005) conducted econometric analysis using institutional level panel data from College Board and IPEDS Faculty Salary Survey. Their research indicate d that the increasing presence of part time faculty or the full time non tenured faculty ha d a negative impact on graduation rates in institutions of higher education. Specifically, a 10% increase in the percentage of part time faculty was associated with
33 2.65% decrease in institutional level graduation rate. Comparably, a 10% increase in the perce ntage of full time non tenured faculty was associated with 2.22% decrease in institution level degree completion. In each occasion, employing a higher percentage of part time faculty has a larger negative impact on student degree completion in public highe r education institutions than in private higher education institutions. However, using institutional level data is not enough to conclude that the degree completion decrease is due to the employment of part time faculty. As the researchers indicated, indiv idual level data that link to part time faculty employment is essential. Umbach (2007) examined the unintended effect of contingent faculty upon student degree or certificate completion He focused on the effect o f behaviors that engage student learning in 132 colleges and universities. Hierarchical linear modeling was used to examine what are the individual and institutional characteristics that contribute to the desirable faculty behaviors that promote effective student lea rning. He found part time faculty were related t o unsatisfying job performance. Calcagno, Bailey, Jenkins, Kienzl, and Leinbah (2008) investigated what institutional factors impact community college student degree or certificate completion They contended the current study failed to account f or institutional factors when investigating community college student academic achievement In particular, they examined whether general institutional charact eristics (i.e. institution size and emplo yment of part time f aculty ), compositional characteristics of students (i.e. percentage of part time students, percentage of female students, and percentage of minority students), financial variables related to revenues and expenditures (i.e. federal student aid per FTE (Ful l Time Equivalent) in state tuition, and average expenditures
34 per FTE in instruction, academic support, student services, and administration), and location characteristics(i.e. urban, suburban, town, or rural) have an impact on student degree completion i n two year community colleges. Their research revealed that or certificate completion than institutional factors are. Jaeger and Eagan (2009) examined the relationship between th e percentage of part time faculty in a community college and California community college system. By using hierarchical linear modeling, the researchers analyzed data on 178,9 85 students in 107 community colleges in California. They found that a 10% increase in overall exposure to part time faculty is associated with 1% decrease in associate degree completion (Jaeger & Eagan, 2009). Student Degree or Certificate Comp letion Research on Student Degree or Certificate Completion Student academic achievement defined as student degree or certificate completion becomes a primary research focus for educational researchers (Tinto, 1975). Although higher educa tion have improved student access to post secondary education, access alone is not enough to ensure student academic achievement (Scott Clayton, 2011). Indeed, policy makers regard retention and graduation rates as indicators for institutional performance. Policy makers at federal level even considering policies that tie eligibility for federal level financial aid programs to institutional graduation rates (Titus, 2004). Researchers have investigated the degree completion of low socioeconomic status (SES) s tudents (Titus, 2006), first generation students (Ishitani, 2 006), students in rural areas (B yun, Irvin, & Meece, 2012), doctoral
35 students ( De Valero, 2001) in four year colleges or universities (Bettinger & Long, 2010; Ehrenberg & Zhang, 2005). In addition, researchers investigated the degree completion of students in two year community colleges (Calcagno, Bailey, Jenkins, Kienzl, & Leinba ch, 2008; Jacoby, 2006 ; Jaeger & Eagan, 2009 ). Theoretical Models Predicting Student Dropout or Retention Researchers hav persistence and degree completion behaviors. Specifically, Student Integration Model, Student Involvement Theory, Causal Model of Student Attribution Student Retention Framework and a n Integrated Model towards Student Persistence are reviewed. in sufficiency in integrating in to society odds of likelihood of suicide in a society. He maintained their decision s to stay in that college (Tinto, 1975). And the interaction between commitment determine whether or not students decide to withdraw from college s Tinto ( 1976) maintained that students who were likely to persist should be both academically and socially engaged in the institutional environment. The match between institutional environment and student commitment to their academic goals was integral for students to have successful persistence. Integration Theory (Astin, 199
36 physical and psycholo gical investment in their college experiences Astin (199 9) articulated the role of student active participation in student learning process. Student highly involved student is depicted as a student who devo tes considerable time in study, be actively involved in student activities, and actively interact with faculty and peer students And an uninvolved student is described as a student who spends little time in study, inactive in student activities, and little or no interactions with his peers or faculty. Student involvement theory realizes that student involvement matters and it has the most impact on first year students (Tinto, 2006). However, both student integratio n behaviors (Hurtado & Carter, 1997 as cited in Titus, 2004). Unlike student integration theory or student involvement theory, Bea attrition model recognize d decision to leave the institution s of higher education. Bean (1980) developed the student attribution m odel based on r esearch work of Price (1977) on turnov er in work organizations, who proposed that organizational characteristics were expected to have a di rect impact on job satisfaction, and job satisfaction is expected to ex ert an influence on employ ee turn over. Similarly, Bean (1980) hypothesized that stud ent dropout was impacted by institutional environment and student satisfaction about their college experiences. Specifically, the model consists of four kinds of variables: dependent variable, dropout; the intervening variables, organizational determinants and the background variables (Bean, 1980). Organizational determinants (i.e. organizational
37 which in turn impact student dropout behaviors. Background variables were h ypothesized to interac t with institutional variables. Built upon the existing literature, Swail (1995) proposed an institutional retention framework which centered on departmental collaborations between recruitment and admissions, academic services, stude nt services, curriculum and instruction, and financial aid services, with student monitoring system being t he center of these servic es (Swail, Redd, & Perna, 2003). Swail maintained that these departments should work towards common goals and emphasize stu system plays a central role in student retention in that it collects systematic data that reveals the true nature of student and faculty life. In sum, Swail and the college colleagues (2003) pointed out that in stitutional retention efforts should be based on collecting and analyzing data, a nd departmental collab orations rather than merely relying on student affairs professionals (Tinto, 2006). Terenzini and Reason (2005) summarized a conceptual framework that i ncorporated both student and organizati onal effects on student persistence The proposed conceptual framework includes four groups of variables that might have an impact on student persistence. The proposed model h ypothesized that students entering in stitu tions of higher education with a set of characteristics that may influence their subsequent interaction s with peers and college environments. The college experience encompasses internal organizational context, peer environment, and eriences
38 Critical Review on Theoretical Models on Student Retention A review of these models indicates that student prior academic preparation, demographic characteristics, student college experience, and institutional environment strongly predict studen While educational researchers proposed various ways in incorporating these variables in their theoretical models, they have limited information regarding how these variables work together & T erenzini, 2005; Reason, 2009). In addition, all of the above mentioned models were developed in four year institutions and were designed for traditional aged, residential stu dents within these institutions of higher education (Bailey & Alfonso, 2005) As such, they the experience of students in two year community colleges (Tinto, 2006) In fact, four year models of attribution have been applied to community coll ege with ambiguous results. The majority of these studies account for 8% to 25% of total variance in attrition, which indicates we still have a long way to go to completely know the reason why students drop out from community colleges (Schuetz, 2005). St udents in community colleges differed drastically from students in traditional four year institutions. Community colleges are open access institutions of higher education which serve a high percentage of part time students, minority students, first generat ion students, non traditional aged students, and low income students. These student characteristics are associated with low graduation rates (Bailey, Calcagno, Jenkins, Kienzl, & Leinbach, 2005 ; Tinto, 2006 ). As such, researchers should be cautious in dire ctly applying these student retention models to investigate the effect of these variables on student degree or certificate completion. As such, the current study serves as an exploratory study to study the effects of variables on student degree or
39 certific ate completion, with an emphasis on the estimation of the percentage of part time faculty upon student degree or certificate completion in two year community colleges. Based on the student retention models reviewed above, three groups of variables are iden year institutions, namely student demographic and prior experiences, student college experience and institutional characteristics. The following conceptual framework is used to predict student degree or certificate completion in two year community colleges. Figure 2 1 : T he conceptual framework predicting student degree or certificate completion in two year community colleges Variables Impacting Student Deg ree or Certificate Completion In this section, I review the individual and institutional factors th at might have an impact on student odds of degree or certificate completion in two year community colleges. Specifically, general institutional factors, stud ent compositional factors, financial factors, and location factors are briefly reviewed. These variables serve as the institutional level variables for the proposed multilevel logistic regression model s In
40 addition, variables that reflect student demograp hics and academic preparation, and student college experiences are included as student level variables for the proposed model s Student Demographics and Prior Experiences Student academic preparation likelihood of g r, race/ethnicity, and age ) (Reason, 2009). Indeed, student academic preparation is one of the strong pr edictors of student academ ic achievement in college. Indeed, a high quality of high school curric than standardized testing such as ACT or SAT scores (Adelman, 2006). Indeed, high schoo l GPA is a much b etter factor year or six (Bowen, Chingos, & McPherson, 2009). Bowen, Chingos, and McPherson (2009) studied 21 public flagship universities and 47 state system universities and fou nd the effect of GPA on student graduation rate is twice to five times as large as the that of standard test scores on student graduation rates. Gender Researchers have long been aware that there is a gap in college degree comp letion rate between male and female students (Buchmann & Diprete, 2006; Brunn, 2009; Goldin, Katz, & Kuziemko, 2006; Jacoby, 1996, 1999; Perna, 2005). The existence of female advantage in college graduation rate is evident among all racial and ethnic group s in United States and most industrialized societies (Buchman & DiPrete, 2006). According to the research conducted by Buchman and Diprete (2006), female students earn more than 60% of all baccalaureate degrees award ed to African
41 Americans, Hispanics, and Native Americans and they earn more than 50% of baccalaureate degrees conferred to Asian Americans and Caucasian Americans. They also reported female advantage in college graduation is prevalent among the majority of European countries and in Australia, Canada, and New Zealand (Eurostat, 2002; Organization for Economic Cooperation and Development [OECD, 2004] as cited in Buchman & DiPrete, 2006). About 61% of female first time, full time students who s year institution i n fall 2004 completed their degr ee at that institution within six years while only 56% of male students did so (NCES, 2012). In 1960, 65% of all bachelor degrees were awarded to male students. Female students continue to lag behind in college completion rate during 1960s and 1970s, until 1982, when female students reached parity with their male counterparts. Since 1982, the percentage of bachelor degree conferred to female students continued to increase to the extent that by the year 2004, female students accounted for 58% bachelor degree recipients (Buchman & DiPrete, 2006). Indeed, female students become the majority of college degree recipients at associate degree (60%), baccalaureate degree (57%), and master degree (59%) levels (Perna, 2005) while man continue to receive the majority of professional (54%) and doctoral degrees (55%). By the year 2009, the percentage of females who had obtained a baccalaureate de gree was eight percent higher than their m ale counterparts (Aud et al 2012). Two factors tha t contribute to this so called female test scores and math and science course taking, and male problems in high schools (Goldin, Kata, & Kuziemko, 2006).
42 Race and ethnicity Early researchers observed that there was a baccalaureate degree attainment gap between African American and Whites. As Kodrzycki (2004) observed since 1970, high school graduation gap between African American students a nd their W hite c ounterparts reduced by 50%, but the bachelor degree attainment gap between these two groups doubl ed. The recent released report Digest of Education Statistics 2011 indicate d between 1971 and 2011, the percentage of students who has achieved a baccalaureat e degree increased from 19% to 39% for Whites, from 7% to 20% for African Americans, and from 5% to 13% for Hispanics. The education achievement gap betw een African American and their W hite counterparts have increased from 12% to 19%, and the gap between H ispanics and their White counterparts have increased from 14% to 26%. In addition, the education achievement gap between Asians/Pacific Islanders and their White counterparts increased from 16% to 18% ( Snyder & Dillow 2012 ). Community colleges are particu larly important for ethnic minorities such as Hispanics, African Americans, and Native Americans. These minority groups are over represented in community colleges. And Hispanic and African Americans are less likely to receive degrees or certificates than t heir White peers ( Bailey, Alfonso, Calcagno, Jenkins, Kienzl, & Leinbach, 2004). Student College Experiences Faculty student interaction Social capital theory informs the study from faculty student interaction perspective. Institutional agents such as fac ulty, administrators, and staff are believed to be essential for student persistence and degree completion (Jaeger & Eagan, 2009). Through these institutional agents, students can get social capital necessary for their academic studies in institutions of h igher education. Social capital is unevenly distributed among students. Students taught by
43 part time faculty are disadvantaged in that students might have less time interacting with their instructors ( Schuster, 2003). Researchers maintained ement degree or certificate completion (Astin, 1999; Benjamin, 2002; Hagedorn et al 2000; Terenzini, Pascarella, & Blimling, 1999 as cited in Chang, 2005; Reason, Terenzini, & Domingo, 2007). Researchers such as R eason, Terenzini, and Domingo (2007), believe d and degree or certificate completion ha d Successful teachers not only possess strong instruction skills and are also avai lable after class. Researchers maintained that there was a positive association between faculty Pascarella, & Blimling, 1996). In addition, faculty student interaction makes students become more satisfied with their college experience (Reason, Trerenzini, & Domingo, 2007). Part time faculty are portrayed as having less interaction with their students than their full time counterparts do. In fact, full time faculty reported spending t wo to four times as many outside class hours as their part time colleagues (Benjamin, 2002). Part time faculty are less likely to participate in faculty student interactions in that part time faculty who account for two thirds of faculty in community colle ges are seldom compensated for office hours and they have courses in multiple campus(Schuetz, 2005). In two year community colleges, a relative lower level of social involvement has been observed. For example, about 20% of students attending two year coll ege participate in school clubs while the percentage of students participating in school clubs reaches between 50% and 60% of students at four year universities or col leges.
44 Compared with their counterparts in four year universities, a high percentage of s tudents are commuter students, attending community colleges on a part time basis and balancing job and family commitments (Chang, 2005). While most research studies on faculty student interaction focus on four year colleges or universities, relatively few research studies are conducted at two year community colleges (Chang, 2005; Hagedorn, Maxwell, Rodriguez, Hocevar & Fillpot, 2000). Hagedo rn, Maxwell, Rodriguez, Hocevar, & Fillpot (2000) studied the peer relations and faculty student relations in a mediu m sized community college on the West Coast and found neither female nor male students were involved in social relations outside the classroom. Chang (2005) used Transfer and Retention of Urban Community College Students (TRUCCS) and found community colleg e students normally have low levels of student engagement with faculty. Students are more likely to interact with faculty on coursework but are less likely to have social involvement with faculty beyond coursework. Unlike their counterparts in four year co lleges or universities, the primary concern of community colleges involves the curricular and academic issues. This is consistent with what the Community College Survey of Student Engagement (CCSSE) 2012 has found out. Built upon existing research, the CCS SE surveys students about their college experiences in community colleges: how they spend their time, their academic gains from classes they have taken; how they evaluate their interaction with faculty members, etc. The CCSSE used cohort data and it includ es 453,093 students from 710 institutions (these students participated in 2010, 2011 and 2012 CCSSE surveys). The survey results indicate d most students in community colleges communicate d with their instructors through email and receive d feedbacks
45 from the ir instructors in tim have much meaningful interactions with instructors after class ( CCSSE 2012). Researchers such as Bailey and Alfonso (2005) questioned the effectiveness of faculty ntion and believed that there are many reasons why students want to interact with faculty (i.e. students and faculty share common values). Tuition and fees. For college students, a ttending community college is significantly cheaper than attending four yea r institutions The average tuition and fees for full time, in state student attending four year public universities are two times higher than the average tu ition and fees for full time in state student attending community colleges. And t he average tuition and fees for full time, in state student attending four year private universities are ten times higher than that for full time, in state student attending community colleges ( Planty et al 2008). Between 1999 2000 and 2006 2007, the average increase in student tuition s and fees for private four year institution are seven times as much as the average increase in student tuition and fees for two year community colleges. Whereas, the average increase in student tuition and fees for public universities is mo re than four times as much as average increase in two year community colleges. The average full time tuition at two year community colleges was about 3,131 dollars for full time student (Baum & Ma, 2012 ) which is the lowest in higher education. But ther e are still many students who struggle to pay for their education (Rose & Hill, 2013). The escalating increase in tuition has led to decreased enrollment in two year
46 community colleges where low income students are over represented (Education Commission of the States, 2003 as cited in Tinto, 2004). Financial aid. While there is no consensus on why a considerable number of constraints. The accelerating cost of a ttendi ng higher education in United States makes financial aid an indispensable tool to increase student access to post secondary education for low and middle income students (Riegg, 2008). According to 2007 2008 National Postsecondary Student Aid Study(NPSAS:0 8), 47.6% of community college students received some aid as compared to 70% of four year college students receive some aid. Although the relationship between financial aid and student retention has been investigated extensively, the impact of financial ai d on student academic success is still unclear (Curs & Harper, 2012). However, the relationship between financial aid and student academic achievement is believed to be positive given the financial aid would allow students to have more time devoted to lear ning rather than spending time working (Alon, 2005 as cited in Curs & Harper, 2012; Dynarski, 2005 as cited in Stater, 2009 ; St. John 1989 ). Community college students rely heavily on federal financial a id programs such as Pell Grant program to pay for th eir education ( Rise & Hill, 2013). Indeed, Pell Grant program plays a major role in providing access to low income students. In the year 2011 2012, 11.2 billion dollars have been awarded to 3.3 million community college students (AACC, 2013). Without finan cial aid, students are more likely to work long er hours to cover the educational cost. Working hours. To pay for the rising price of higher education, a higher percentage of students choose to work while attending institutions of higher education
47 (Nonis & Hudson, 2010; Riggert, Boyle, Petrosko, Ash, & Rude Parkin s, 2006). In addition to financing their higher education, many students choose to work during their academic studies in that they can gain valuable working experiences (Light, 2001). Inevitably, student academic studies will be impacted when students devote much time to non academic related work. Indeed, researchers have long been interested in investigating the relationship between student working hour s and student learning outcomes (i.e. GPA). H owever, they presented a really mixed and sometimes contradictory picture about the effect of student working hours experience (i.e. GPA, degree completion) (Riggert, Boyle, Petrosko, Ash, & Rude Parkins, 2006). Austin (1975) found o n campus employment can improve student academic into the campus environment. Working part time seemed to have a positive impact on g hours are fewer than 15 20 hours per week Dundes and Marx (2006) reported working part time forced students to become more efficient in their academic studies. In addition, students tend to reduce leisure time rather than time devoted to academic studie s (Fjortoft, 1995 as cited in Cheng & Alcantara, 2004; Tyson, 2011). Pascarella, Bohr, Nora, Desler, and Zusman (1994 as cited in Cheng & Alcantara, 2004) reported working more than ten hours on campus Terenzi, Yaeger, Pascarella, and Nora (1996 as cited in Cheng & Alcantara, 2004) found work study program slightly
48 Terenzini (2005) found there was no direct evidence be tween student employment and Researchers are aware that student working hour s will not only have an impact on student college experience but student degree completion. Hanniford and Sagaria (1994) examined the influence of employment and family commitments on student associate or baccalaureate degree completion. Two samples were drawn from National Longitudinal Study of the High School Class 1972: associate and baccalaureate degree seekers. And their research indicated that working full time ha d a negative impact on baccalaureate degree seeking males and associate degree seeking females. Riggert and the colleagues (2006) found working more than 20 hours significantly contribute to g less than 20 hours seemed to have no Financial loans. Using data from the 1987 National Post secondary Student Aid Survey, St. John and Paulsen (2001) found student loans had a negative effect on both poor and wor king class students when investigating persistence rates among undergraduate students. Studying a single institution data, DesJardins, Ahlburg, and McCall (2002) found student persistence to highe r education. According to 2007 2008 National Postsecondary Student Aid Stud y (NPSAS:08), 13.2% of community college students took some loans as compared to 47.8% of four year college students do so. Student loans have indirect impact on student degree atta inment (Dowd, 2008; St. John & Paulsen, 2002). Dowd (2008) pointed out low effect on their degree completion. According to the researcher, low income students
49 were reluctant to borrow money and therefore might miss opportunities of attending elite universities, which leads to a rewarding career. Li (2008) drew data from the Beginning Postsecondary Student Longitudinal Study (96/01) survey and employed multilevel analysis to examine the different completion. Li (2008) maintained that students who used grants as the only means to finance their higher education were 50% more likely to obtain their degree than those who used only loans. In addition, stud ents who received both grants and student loans need more time to complete their college degree. Therefore, student loans are not as effective as grants in helping students persist in their academic studies and get their degree s (Gladieux, 2002). Given li ttle research focused on the effects of student loans on community college degree completion, Dowd and Coury (2006) informed the educational research persistence to th colleges. They nevertheless cautioned that the findings were confounded by high uncertainty of degree attainment in community colleges (many community college students do not plan to com plete a bachelor degree). On the other hand, research on student dropout confirmed the negative effect of student loans on student educational achievement. For instance, Gladieux and Perna (2005) found students who borrowed and enrolled in four year colle ges in 1995 1996, 19% had dropped out. Those w ho borrowed and enrolled in two year community colleges 24% dropped out. Those who borrowed and enrolled in private for profit yet less than four year institutions, 32%
50 dropped out. Their report indicated that a large number of student borrowers dropped out regardless what type of post secondary institutions they had enrolled. However, some researchers pointed out the positive impact of student loans on degree achievement Stampen and Cabrera (1988) found all forms of financial aid packages effective in enabling low income students to persist in higher education at rates equal to higher income students who had not received any aid. Loans proved to be an effective way of financing higher education for most stude nts (Gladieux & Perna, 2005; Lam, 1999). In fact, student loans might make students highly motivated to complete their degree because the debt they have accumulated during their college years should be paid back (Lam, 1999). Gladieux and Perna (2005) indic ated student loans might prove to be a better alternative compared to working part time, which may or certificate completion. Others cautioned although student loans might be less expensive for government to fina nce higher education, their impact on loan recipients are more complicated (Long, 2008; Pascarella & Terenzini, 2005). Singell (2002) found subsidized loans had a insignificant ef student loans: while the availability of loans might widen opportunity for some students, they college degre e s Institutional Characteristics General institutional factors These factors include institution size, the percentage of part time faculty. As indicated above, the employment of part time faculty on student achievement is inclusive. Critics claimed th at hiring a high percentage of
51 part time faculty has a negative impact on student degree or certificate completion Having a high percentage of part time faculty may have a negative impact on the construction of supportive environment for students ( Calcagn o, Bailey, Jenkins, Kienzl, & Leinbach, 2008 ). On the other hand, advocates argue that hiring a high percentage of part time faculty has no or positive impact on student academic achievement The effect of institution size is also inclusive. Kamens (1971 a s cited in Titus, 2004) maintained increases the persistence by four percent Yet, other researchers argued that a relatively smaller institution might create a more socially and academically integrated environment which is particular true for community colleges (Bailey, Calcagno, Jenkins, Kienzl, & Le inbach, 2005; Pascarella & Terenzini, 1991, 2005). Student compositional factors. Student compositional variables include the percent age of female students and minority students, and the percentage of part time students. Female students account for 57% of students at two year community colleges in 2010 and currently more than 4 million female students are enrolled in these two year institutions ( Rose & Hill, 2013). In addition, m inority students account for a large percentage of student body in two year co mmunity colleges. They account for 47% of student body in two year community colleges and account for 39% in public four year institutions (Rose & Hill, 2013). In community colleges, part time students account for 64% of student body in community colleges and 29% of student body in four year colleges and universities (Bailey & Alfonso, 2005). Having a high percentage of part time students and minority students would have a negative impact on student
52 engagement and integration in community colleges ( Calcag no, Bailey, Jenkins, Kienzl, & Leinbach, 2008 ). Research studies on selective four year colleges indicate d that students working with high performing students tend to benefit. Since female students are high performing students, a high percentage of femal e students in community colleges might as cited in Calcagno, Bailey, Jenkins, Kienzl, & Leinbach, 2008 ). However, research report written by Rose and Hill (2013 ) indicated that nearly 25% of female students in community colleges ha d significant commitments besides academic studies (i.e. work and/or family responsibilities). These students had limited resources and were unprepared for the academic studies in communit y college s (Rose & Hill, 2013), which may place them in a disadvantaged position in degree or certificate completion. Financial factors. Institutional commitment includes average in state tuition, average expenditure for instruction per FTE, average expenditure fo r academic support per FTE, average expenditure for student services per FTE, and average expenditure f or institutional support per FTE As indicated by the Student Retention Framework proposed by Swail, Redd, and Perna (2003), the average expenditure for instruction, academic support, and student services promote student retention. This is consistent expenditures on six year cohort graduation rate at 363 four year baccala ureate institutions, the research er reported there was a significant positive relationship between instructional and academic support expenditures and cohort degree completion rate. According to Titus (2004) and Ryan (2004), expenditure for
53 administrative services has a negative impact on student degree completion in four year colleges or universities. Similarly, expenditure for institutional expenses in two year community colleges might be negatively rel ated to student degree or certificate completion. In addition, the average in state tuition is believed to have an impact on student degree or certificate completion (Calcagno, Bailey, Jenkins, Kienzl, & Leinbach, 2008). The Federal level Pell Grants provided need based financial assistance to promote studen provided by Ohio Board of Regents (OBR), Bettinger (2004) investigated the effect of Pell grants on student dropout behaviors and found it was effective in retaining students in insti tutions of higher education. Fixed location factors. The fixed location factors include only one variable: the level of U rbanicity. It indicates whether a two year community college is located in rural, town, suburban or urban area Chapter Summary The s ection briefly reviews the literature pertain ing to current study. The history of community college s provides the context of institutions of higher education under investigation. Community colleges are unique post secondary education institutions that prim arily serve non traditional students with an emphasis on associate degree and or certificate completion, academic tra nsfer from community colleges to four year universities, workforce development, and remedial education. Research studies pertain ing to part time faculty in community colleges are thoroughly reviewed. Student retention and persistence models are critically investigated and a student retention framework for community college students is presented based on thorough litera ture review. Variables p ertain ing to the proposed student retention framework are reviewed.
54 CHAPTER 3 METHODOLOGY The impact of part time faculty on student degree or certificate completion has been investigated using qualitative and quantitative approaches ( Benjamin, 2002; Betti nger & Long, 2010; Calcagno, Bailey, Jenkins, Kienzl, & Leinbah, 2008; Cohen & Brawer, 2003; Curtis & Jacobe, 2006; Eagan & Jaeger, 2008; Ehrenberg & Zhang, 2005; Jacoby, 2006; Jaeger & Eagan, 2009; Sandy, Gonzalez, & Hilmer, 2005; Umbach, 2007 ) While the qualitative research provides the theoretical and empirical foundations for unders tanding part quantitative research examines whether there is an effect in a population and the extent of such effect. Researchers have long been aware of the hierarchical nature of the data and have examined ways to incorporate both individual and group level information into regression analysis. Approaches usually inc lude either aggregating student level variables to the instituti on level or disaggrega ting institution level variables in to student level (Leeuw & Meijer, 2 008). However, both approaches appear to be problematic as cross level individuals nested in the same group are more similar to each other than the individuals not nested in the same gr oup (Hoffman & Gavin 1998). Indeed, estimating the effect of part tim e faculty academic achievement using ordinal least square regression analysis raises the question of unit of analysis: whether we should regard student or institution as the unit of analysis. If the collected data are analyzed at the student level, which means we ignore the nested structure of individuals within organizational units, the estimated standard errors will be too small and T ype I errors will be inflated (Raudenbush & Bryk, 2002; Umbach & Wawrzynski, 2005). On the other hand, if the data are analyzed at the
55 institutional level, it is questionable to include other student level variables into the multiple regression model s The key advantage of mu ltilevel modeling over the traditional ordinal least square regression modeling model can represent these features (Raudenbus h & Bryk, 2002). The purpose of the current study is to investigate the effect of part time faculty three pertinent research questions are raised and answered using the proposed two level logistic regression models : 1. Whether part time faculty significantly impact student degree or certificate completion while holding individual characteristics and other institutional level variables constant. If there is a significant eff ect on student degree or certificate completion, the magnitude and direction of such effect. 2. Whether individual level variables and other institutional variables significantly impact student degree or certificate completion. If there are significant effec ts, the magnitude and the direction of these effects 3. Are there any difference s in terms of variables that impact student degree or certificate completion Research Design, Datasets, and Sample Size The curren t research adopts an ex post facto research design by selecting a nationally representat ive sample of community college students and follows them from the academic yea r 2003/ 04 to the year 2008/ 09. The 20 04/09 Beginning Postsecondary Students Longitudinal Study (BPS: 04/09) was conducted by U.S. Department of ed data on topics such as student demographic characteristics, school and work characteristics, and student persistence and deg ree attainment during three time points: their first year, third year, and sixth year in post secondary education institutions The target population
56 included a sample of students who started their post secondary education for the first time during the 200 3/ 04 academic year at any institutions of higher edu cation in United States that were eligible for National Postsecondary Student Aid Survey(NPSA:04). Th e final dataset has enough data information for 16,680 students (Wine Janson, Wheeless, & Hunt White, 2011 ) The Integrated Postsecondary Education Data System (IPEDS) is a system of interrelated surveys conducted by NCES. The IPEDS collect ed information from institutions of high er education that participate the federal financial aid p rograms. The IPEDS d ata include d information about college enrollment, program completions, graduation rates, faculty and staff, fi n ances institutional prices, and student financial aid (IPEDS, 2013). The IPEDS forms the sampling frame for the NCES surveys such as National P ostsecondary Student Aid Survey. As such, respondents in BPS a re nested in the institutions sampled by IPEDS and it is reasonable and meaningful to merge the two datasets. The student l evel data are drawn from 2003/ 04 BPS dataset as students entered the i nstitutions of higher education first time. And the institutional level data are drawn from IPEDS. I merged the IPEDS dataset and BPS dataset by using the shared institution ID ( community college institution ID). The final merged dataset contain s 19 40 obse rvations for 5 0 community colleges. The sampling weight s for BPS dataset are normalized and are taken into account when conducting the multilevel logistic regression analysis using HLM 7.0 R code s for merging the datasets and normalizing the sampling weig hts are presented in the appendix section ( APPENDIX B )
57 Data Source and Variables As indicated above, data used in this study are drawn from Integrated Postsecondary Education Data System (IPEDS) and 2004 /2009 Beginning Postsecondary Students Longitudina l Study (BPS: 04/09). Specifically, data pertain ing to institutional characteristics are drawn from IPEDS which include data information about percentage of part time faculty, institutional enrollment, financial aid, and aggregated student characteristics. Specifically, four subsets of institutional characteristics are drawn from the dataset. The first subset is general institutiona l characteristics which include institution size, percentage of part time faculty. The second subset is student compositional c haracteristics, which consist of the percent age of part time students, the percent age of female students, and the percent age of minority students. The third subset, financial characteristics include the federal student aid per FTE ( Full time equivalent ), t he average undergraduate in state tuition, the academic support per FTE, the student services per FTE, and the administration expenses per FTE. The fourth subset, fixed location characteristics include the geographical location of a community college The criteria for selecting the participant community colleges are higher education institutions that are Title IV participating institutions in the United States (all the states within the United States are included). The higher education sectors include publ ic, two year; private not for profit, two year; and private for profit, two year institutions. The highest degrees offered are associate and professional certificates. All of these two year community colleges are categorized as associate colleges. Data pe rtain ing to student level variables are drawn from The Beginning Postsecondary Students Longitudinal Study (BPS: 04/09) dataset. The BPS (04/09)
58 dataset follow ed a cohort of students who were enrolled in post secondary education for the first time. The sur vey collect ed data on student persistence, degree or certificate completion as well as other indicators for student learning outcomes. Students of the most recent cohort of BPS, BPS: 04/09, were first surveyed at the end of their f irst academic year (2003 / 04). The first follow up interview (BPS: 04/06) was administered to their entering institutions of higher education with a particular focus on topics such as continued educ ation experience, education fina ncing, and entry into workforce. The second follow up interview (BPS: 04/09) was academic progress in six academic years after their entering institutions of hi gher education with a special focus on degree completion (Wine, Janson, Wheeless, & Hunt White, 2011) Dependent variable, PRAT3Y (Persistence and attainment anywhere from the year 20 03/ 0 4 to the year 20 05/ 06) indicates whether the respondent attained a ny certificate or degree, or was still enrolled at any post secondary institution as of June 2006. Dependent variable, PRATT6Y (Persistence and attainment anywhere from the year 20 03/ 0 4 to the year 2008/ 09) indicates whether the respondent attained any certificate or degree or was still enrolled at any post secondary institution as of June 2009. Student level variables include GENDER, RACE, HCGPAREP (High School GPA), TEACTDER (Derived ACT Score), TESATDER (Derived SAT Score), TOTAID (Total Aid Excluding Work Study), JOBHOUR (Hours work per week) and indicate the average hours the resp ondent work every week; TUITION (Tuition and fees in 2003/ 04)
59 attended during the 2003/ 04 academic year; ATTEND (Attend ance intensity in fall 2003 / time student FREQ A (Faculty informal meeting) indicates whether or how often the respondent had informal or social interaction with faculty mem bers outside of classro om and office during the 2003/ 04 academic year. FREQ B ( Faculty talk outside class) indicates whether or how often the respondent talked with faculty about academic m atters outside of the classroom including email durin g the 2003/ 04 a cademic year. Multiple Imputation Missing values are a common issue in complex survey data. And se veral traditional methods (i.e. list wise deletion, pairwise deletion, mean substitution, regression based single imputation ) have been implemented to addres s this issue (Peugh & Enders, 2004). Yet, these traditional m ethods have been criticized as they produce biased sample statistics (Peugh & Enders, 2004). New method such as m ultiple imputation is developed to handle missing values in the datasets and has b een considered as the recommend procedures in the research literature ( Peugh & Enders, 2004; Rubin, 1987 ).The method is particularly useful in complex surveys where non responses for survey questions are common (Little & Rubin, 2002). Multiple imputation r equires data missing at random (missing values on a particular variable can be associated with other measured variables but must be uncorrelated with underlying values of the variable) or missing completely at random (missing values on a certain variable a re unrelated to other variables and underlying values of the variable itself) (Peugh & Enders, 2004). In this study, multiple imputation is used to account for the missing values in the survey data. Compare with other methods that deal with missing values in complex
60 surveys (i.e. list wise deletion, pairwise deletion, mean substitution, regression based and single imputation), multiple imputation yields unbiased parameter estimates and standard errors. Multiple imputation produces multiple datasets and rep laces the missing values with two or more values that denote a distribution of possibilities (Rubin, 1987). Then, statistical analysis should be performed with each of these imputed datasets, which a set of analytical results for each dataset. These result s can be combined to yield an inference (i.e. which is usually the average of estimates from multiple imputed datasets) that reflects sampling variability due to the missing values (Little & Rubin, 2002). (3 1) The within imputation variance (average of the squared standard errors): ( 3 2) The between imputation variance (variance of the estimates): (3 3) The total variance (the standard error is the square root of T) (3 4) In the current study, I use statistical software R to produce multiple imputed datasets with its MICE package ( A PPENDIX B ). Given there has been a modest fraction
61 of informa tion missing, the number of imputed datasets should between two and ten (Rubin, 1987). In this study, I create five imputed datasets and these imputed datasets are input i nto HLM7.0 to produce an aggregated parameter estimates and standard errors Logisti c Regression Since the dependent variable is dichotomous, logistic regression is used to address assumption violations: equal conditional variance and conditional normality. The log odds form is presented as follows: (3 5) Where is the intercept and it is the log odds when the independent var iables are held constant at zero is the amount the log odds change when X changes one unit. The term refers to residu al error The advantage of this format for the model is that we can have the interpretations that are s imilar to those in multiple regression analysis Multi level Modeling: Concepts and Application Multilevel modeling can be illustrated by using the two level hierarchical linear model. At the student level (within community colleges ) the units are students and each student outcome (i.e. degree or certificate completion) can be formulated as a function of individual characteristics. At the institutional level (between community colleges ) the units are organizations. The regression coefficients and intercepts are regarded as outcome variables that are impacted by organizational level characteristics. Multi level modeling provides a solu tio n to distinguish between within school and between school variances.
62 Multi level analysis involves two steps: unconditional and conditional analysis. The first step is to perform the un independent variabl es in the model. The unconditional model provides preliminary information regardin g the within school and between school variance. If the unconditi onal analysis shows no significant variance across community colle ges, the second step is not necessary Othe rwise, if the unconditional analysis indicates significant variation across community colleges, the second step should be carried out Student level model (level 1). We denote the outcome for person in organization as The outcome is represented as a function of individual characteristics, and a model error : (3 6) Institutional level model (level 2). The regression coefficients and intercept are assumed to vary across units. This variation is in turn modeled in a set of Q +1 level 2 functions : one for each of the regression coefficients and intercep t from the level 1 model. Each is modeled a s an dependent variable that is impacted by a set of organization level variables, and a unique organization effect, Each has a model of the form (3 7) The coefficients represent the influence of organizational variables, on the within organization relationships represented by We assume that the set of Q+1
63 level 2 random effects is multivariat e normally distributed. Each has a mean of 0, some variance, and with covariance, In the current study, Intercepts as Outcomes m odel (IAO) model is proposed to fit the collected data The model assumes that the intercepts are random and vary across different community colleges But the model assumes that the slopes are fixed as the sample size for the level 2 is small Student level variables are hypothesized to predict indiv certif of completing their degree or certificate is hypothesized to vary across different odds of degree or certificate co mpletion in two year community colleges Given the dependent variable is a binary variable; the proposed multilevel model is specified as a two level logistic regression model The two level logistic modeling is presented as follows: Level 1 Model: Prob( PRATT3Y ij /PRATT6Yij=1| j ) = ij (3 8) log[ ij /(1 ij ij (3 9) ij = 0j + 1j *( GENDER ij ) + 2j *( RACECEN ij ) + 3j *( TOTAID ij ) + 4j *( JOBHOUR ij ) + 5j *( ATTEND ij ) + 6j *( FREQ A ij ) + 7j *( FREQ B ij ) + 8j *( HCGPAREP ij ) + 9j *( TEACTDER ij ) + 10j *( TESATDER ij ) + 11j *( TUITION ij ) (3 10) Level 2 Model: 0j = 00 + 0 1 *( PROPPTFA j ) + 02 *( INSTITUT j ) + 03 *( URBAN j ) + 04 *( PCTMINOR j ) + 05 *( PCTWOMEN j ) + 06 *( TUITIONF j ) + 07 (PELLGRAN j ) + 08 (PROPORTI j ) 09 (INSTRUCT j ) + u 0j 1j = 10 2j = 20 3j = 30 4j = 40
64 5j = 50 6j = 60 7j = 70 8j = 80 9j = 90 10j = 100 11j = 110 (3 11) The combined model: ij = 00 + 01 *( PROPPTFA j ) + 02 *( INSTITUT j ) + 03 *( URBAN j ) + 04 ( PCTMINOR j ) + 05 *( PCTWOMEN j ) + 06 *( TUITIONF j ) + 07 *( PELLGRAN j ) + 08 *( PROPORTI j ) + 09 *( INSTRUCT j ) + 10 *( GENDER ij ) + 20 *( RACECEN ij ) + 30 *( TOTAID ij ) + 40 *( JOBHOUR ij ) + 50 *( ATTEND ij ) + 60 *( FREQA ij ) + 70 *( FREQB ij ) + 80 *( HCGPAREP ij ) + 90 ( TEACTDER ij ) + 100 *( TESATDER ij ) + 110 *( TUITION ij ) + u 0j (3 12) Centering Option Centering student level variables (i.e. non centering, grand mean centering, and gr oup mean centering) has implications for how researchers interpret the intercept terms. With non centering, the intercept is often meaningless in the context of social science but grand or group mean centering often makes the intercept interpretable. It me ans the expected value of dependent variable when the independent variable is equal to the grand mean or group mean (Hoffman, 1998; Landeghem, Onghena, Van Damme, & Opdenakker, 1999). Although centering issue serves as an important tool in multilevel model ing, there are no guidelines and rules in regards to model centering. If an IAO model is used to conduct multi researchers use non centering, grand mean centering or group mean centering approach. As such, non cent ering approach is used in the current study (non centering and grand mean centering lead to statistically equivalent models). Any estimation, hypothesis test or confidence interva l that can be obtained with grand mean centering approach can also be obtai ned with non centering ap proach
65 Weighting Issue To analyze the cases wh o were research participants in all three of the base year study (NPSAS: 04), the first follow up study (BPS: 04/06), and the second follow study (BPS: 04/09), a weight is constructed in BPS: 04/09. And the weight is called longitudinal or panel weight. This is to ensure the consistency with BPS : 04/09 study weight s and BPS: 04/06 and NPSAS: 04 weights. In the study, WTB 000 is used since it is the response adjusted, calibrated panel w eigh t for the BPS: 04/09. Normalized sampling weights for selected participants are used when conducting multilevel logistic regression analysis using HLM 7.0 software. Limitations of the Study One limitation is that the 2008/09 IPEDS data employed Carneg ie Classification of Institutions of Higher Education 2000 The new version, The Carnegie Classification of Institutions of Higher Education 2005 includes changes in this former version (i.e. geographic location of community colleges and institution size). Researchers should be aware of these changes when interpreting the analytical results of the current study (Carnegie Foundation for the Advancement of Teaching, 2013). The second limitation is that the current research is unable to track community colleg e students transferring to four year institutions without completing an associate degree or certificate. degree or certificate completion in the current study. Since both the IPEDS and BPS are seco ndary datasets, the current study is also limited by the variables and definitions available in these datasets (Jaeger & Eagan, 2009). More variables pertain ing to student characteristics and college experiences, as well as inst itutional characteristics co uld be included in the multilevel logistic regression
66 models. This is evidenced by the analytical results shown in chapter four (Table4 8 and Table 4 9). The significant chi squares indicate that more independent variables should be added in to the analytic al models. Chapter Summary In this methodology section, an ex post facto research design is presented to address the proposed research questions. The datasets utilized in the current study are IPEDS and BPS datasets produced by National Center for Educati on Statistics (NCES). Multiple imputation procedure is used to address missing values in the selected datasets. A total of five imputed datasets are created. The nested data structure justifies the usage of multilevel logistic regression. The proposed Inte rcepts as Outcome (IAO ) m odels justify the usage of non centering approach. Sampling weights are normalized before being input in the analytical mode l s The limitations of the current study are briefly summarized.
67 CHAPTER 4 RESULTS The descriptive statis tics are reported for both dependent and independent variables included in t hese multilevel logistic regression model s Then, unconditional and conditional analyses are presented to illustrate which variables are significant predictors for the dependent va completion in community colleges within three or six years. As indicated by Table 4 1, tificate completion between the year 20 03/04 and the year 20 05/06 ( PRATT3Y or cer tificate completion between the year 20 03/04 and the year 20 08/09 ( PRATT6Y ). Institutional level independent variables are drawn from IPEDS dataset while student level independent variables a re drawn from BPS dataset. Table 4 1: D ependent and independent variables used in the analysis Variable Names Abbreviation Scale points and data sources Dependent Variables Degree or certificate completion between 03/ 04 05/ 06 PRATT3Y Categorical, Not graduate=0, Graduate =1 Source: BPS Degree or certificate completion between 03/ 04 08/ 09 PRATT6Y Categorical, Not graduate=0, Graduate =1 Source: BPS Independent Variables (Student Level) Gender GENDER Categorical, Male=0, Female=1 Source: BPS Rac e RACECEN Categorical, White=0, Minorities =1 Total aid excluding work study TOTAID Continuous, Source: BPS Hours worked per week (including work study) JOBHOUR Continuous, Source: BPS
68 Table 4 1. Continued Variable Names Abbreviation Scale point s and data sources Attending status ATTEND Categorical, Not Full time =0, Full time=1 Source: BPS Faculty informal meeting FREQ A Categorical, Not often=0, Often=1 Source: BPS Faculty talk outside class FREQ B Categorical, Not often=0, Often=1 Sou rce: BPS High school GPA HCGPAREP Continuous 0.5 0.9 (D to D)=1, 1.0 1.4 (D to C )=2, 1.5 1.9 (C to C)=3, 2.0 2.4 (C to B )=4, 2.5 2 .9 (B to B)=5, 3.0 3.4 (B to A )=6, 3.5 4.0 (A to A)=7 Source: BPS ACT score TEACTDER Continuous, Source: BPS SA T score TESATDER Continuous, Source: BPS Tuition and fees TUITION Continuous, Source: BPS Independent Variables (Institutional Level) The percentage of part time faculty PROPPTFA Continuous, Source: IPEDS Institution size INSTITUT Categorical, 0 9,999=0 10 ,000 20,000 and above=1 Source: IPEDS Geographical location URBAN Categorical, Rural =0 Non rural=1 Source: IPEDS Percentage of minority students PCTMINOR Continuous, Source: IPEDS Percentage of female students PCTWOMEN Continuous, Sour ce: IPEDS Tuition and fees(institution) TUITIONF Continuous, Source: IPEDS Pell Grant PELLGRAN Continuous, Source: IPEDS Percentage of part time students PROPORTI Continuous, Source: IPEDS Instructional cost INSTRUCT Continuous, Source: IPEDS Statis tical Software: R, HLM, SPSS R software is used to merge the IPEDS and BPS: 04/09 datasets. In addition, it is also used to perform multiple imputation to account for missing values in certain variables. I use the HLM 7.0 program to produce parameter estim ates for student and institutional level variables. The HLM 7.0 program can be used to fit the models to the dependent variables that generate a linear model with independent variables that
69 account for variations at both individual and organizational level (SSI, 2013). HLM 7.0 program not only estimates model coeffic ients at each level, but also predicts the random effec ts of student and institutional levels. In the current study, student level and institutional level da ta are input first in IBM SPSS 21. An d the n, the HLM 7.0 program utilizes the imputing IBM SPSS 21 files to estimate both fixed and random effects. Descriptive Statistics Descriptive statistics pertain ing to the means and standard deviations of variables are presented in the following two ta bles (Table 4 2 and Table 4 3) : Table 4 2 : D escriptive statistics for level 1(student level) variables N Mean/ Count Std/ Percentage D ependent Variables PRATT3Y 1940 402 20.2% PRATT6Y 19 40 761 38.4% Independent Variables(student level) G ENDER 19 40 1176 59.3% RACECEN 19 40 648 32.7% TOTAID 19 40 2191.40 2891.74 JOBHOUR 19 40 21.48 16.10 ATTEND 19 40 1177 59.3% FREQ A 19 40 88 4.4% FREQ B 19 40 216 10.9% HCGPAREP 19 40 5.31 1.26 TEACTDER 19 40 18.75 4.08 TESATDER 19 40 887.63 172.85 TUITION 19 40 1458.86 1140.87 *The descriptive statistics are calculated based on the five imputed datasets. The formula to obtain means and standard errors of varia bles are based on Rubin (1987 research work on imputed datasets. Ta ble 4 2 and Table 4 3 prese nt descriptive statistics for the student and institutional level variables included in the two level logistic regression anal ysis. As
70 indicated by the above mentioned tables ( Ta ble 4 2 and Table 4 3), means and standard deviation s of independent and depen dent variables are presented. Table 4 3: Descriptive statistics for the level 2(institutional level) variables N Mean/ Count Std / Percentage Independent Variables(institutional level) PROPPTFA 5 0 48.86 27.31 INSTITUT 5 0 865 43.6% URBAN 5 0 1117 56.3% PCTMINOR 5 0 39.18 21.13 PCTWOMEN 5 0 61.29 5.65 TUITIONF 5 0 2550.49 1024.69 PELLGRANT 5 0 2959.37 422.333 PROPPTS 5 0 50.73 13.63 INSTRUCTION EXP 5 0 4959.78 2483.26 *The descriptive statistics are calculated based on the five imputed datasets. The formul a to obtain means and standard errors of variables are based on Rubin (1987 research work on imputed datasets. In the analytic sample, there are a total of 19 40 students nested in 5 0 two year community colleges. As revealed by the descriptive statisti cs, the average financial aid participants received were $ 2191.40 dollars and the average working hours for participants were 21.48 hours. The average percentage of part time faculty in selected communit y colleges was 48.86%. On ave rage, minority students accounted for 39.18% of student population and female students accounted for 61.29% of student population in selected community colleges. Unconditional Analysis ( Random ANOVA ) Hypothesis test should b e firstly conducted to see whether the estimated value of group variability is significantly different from zero (Raudenbush & Bryk, 2002). In
71 the current study, r andom ANOVA (un conditional analysis) is condu cted to assess institutional varia nce. If institutional level variance is in significant, there is no need to conduct multi level logistic regression analysis (conditional analysis) On the other hand, if institutional variance is significant, there is a need to conduct multi level logistic regression analysis. Table 4 4: F inal estimation of variance components for pratt3yij (unconditional) Random Effect Standard Deviation Variance Component d.f. 2 p value INTRCPT1, u 0 0.60 0.36 50 167.96 0.000 imputed datasets. Table 4 5: F inal estimation of variance components for pratt6yij (unconditional) Random Effect Standard D eviation Variance Component d.f. 2 p value INTRCPT1, u 0 0.36 0.13 50 109.16 0.000 *The estimates are calculated in HLM 7.0 multiple imputed datasets. The results of the unconditional analysis are p resented in Table 4 4 and Table 4 5 As indicated by the above mentioned tables the chi squares for random effect s are significant ( =167.96 p<.05; =109.16 p<.05). The significance of chi squares indicate that variances for random effect va ry across community colleges ( Int ra class C orrelation Coefficient (ICC) indicate that degree or certificate completion is between community colleges when analyzing the data between 03/04 and 0 5/06; while 11.58 % of var of degree or certificate completion is between community colleges when analyzing the data between the year 20 03/04 and the year 20 08/09) M ulti level logistic regression analysis is needed to i nvestigate the effect of part time faculty on student degree or certificate completion. Thus, it is important to identify predictors in the institutional level
72 that account for the variance. As indicated above, Intercepts as Outcome model (IAO) is employed to examine whether iden tified variables explain institutional level variability The institutional analytical results are presented in Table 4 6 and Table 4 7 Conditional Analysis Student and institutional level independent variables are then added to the unconditional model s mention ed above. This statistical procedure is known as the conditional analysis as compared to above mentioned unconditional analysis. As indic ated by Table 4 6, multi level logistic regression analysis indicate that the percentage of part time faculty, the phys ical location of the community colleges, the percentage of female students, total aid excluding work, working hours, high school GPA, and tuition and fees odds of deg ree or certificate completion b etween the year 20 03/04 and the year 20 05/06. Table 4 6. Final estimation of fixed effects Fixed Effect Coefficient Standard error t ratio Approx. d.f. p value For INTRCPT1, 0 INTRCPT2, 00 1.74 1.569 1.108 4 0 0.274 PROPPTFA, 01 0.006 0.003 2.020 4 0 0.050* INSTITUT, 02 0.588 0.240 2.450 4 0 0.019* URBAN, 03 0.611 0.191 3.191 4 0 0.003* PCTMINOR, 04 0.007 0.004 0.159 4 0 0.874 PCTWOMEN, 05 0.042 0.022 1.9 55 4 0 0.057 TUITIONF, 06 0.000004 0.0001 0.027 4 0 0.979 PELLGRAN, 07 0.0001 0.0002 0.541 4 0 0.591 PROPORTI, 08 0.002 0.007 0.248 4 0 0.805 INSTRUCT, 09 0.00004 0.00003 1.385 4 0 0.173 INTRCPT2, 10 0.158 0.136 1.166 187 0 0.244 For MINORITY slope, 2 INTRCPT2, 20 0.273 0.175 1.558 187 0 0.119 For TOTAID slope, 3 INTRCPT2, 30 0.00003 0.00002 1.137 187 0 0.256 For JOBHOUR2 slope, 4 INTRCPT2, 40 0.012 0.004 2.729 187 0 0.006*
73 Table 4 6. Continued Fixed Effect Coefficient Standard error t ratio Approx. d.f. p value For FULLTIME slope, 5 INTRCPT2, 50 0.816 0.158 5.157 187 0 0.000* For MEET slope, 6 INTRCPT2, 60 0.128 0.335 0.383 187 0 0.702 For TALK slope, 7 INTRCPT2, 70 0.242 0.207 1.168 187 0 0.243 For HCGPAREP slope, 8 INTRCPT2, 80 0.140 0.070 1.990 3 0 0.050* For TEACTDER slope, 9 INTRCPT2, 90 0.025 0.125 0.203 2 0 0.841 For TESATDER slope, 10 INTRCPT2, 100 0.00007 0.003 0. 024 30 0.981 For TUITION2 slope, 11 INTRCPT2, 110 0.0001 0.00006 2.011 12 30 0.045* P values (p<.05) with indicate variables that are significant predictors for the outcome h work on multiple imputed datasets. Variables pertain ing to student financial loans, average expenditure for academic support per FTE, average expenditure for student services per FTE and average expenditure for institutional support per FET are not inclu ded becaus e of extremely low responses Part time faculty have no negative impact on student degree or certificate attainment. A unit increase in the percentage of part time faculty results in a 0.6% increase in cate completion. Institution size is negatively attend a larger two year community college are 58.8% less likely to achieve degree or certificate completion than students at tending a smaller institution holding other variables constant. Students who pursue their academic studies in suburban, town, city areas are 61% more likely to achieve degree or certificate completion than community colle ge students who are enrolled in com munity colleges located in rural areas between the year 2003/04 and the year 2005/06 High school GPA also has a significant impact on student degree or certificate completion. A unit increase in student high school GPA
74 leads to 14% increase in student de gree or certificate completion. It indicates student academic preparation plays an important role in predicting student academic achievement in community colleges. In addition, attendance status is strongly associated with student academic achievement. Stu dents who are part time students are 81.6% less likely to graduate than their full time counterparts. Working hours are also related to student degree or certificate completion. A unit increase in student ikelihood of degree or certificate attainment Student tuition has a really small impact on student academic achievement and a unit increase in tuition le degree or certificate completion. All other student and inst itutional level variables ha ve no impact on student degree or certificate completion in two year community colleges. When analyzing the data between the year 20 03/04 and the year 20 08/09 (Table 4 7), multi level logistic regression analysis suggests that institution size, percentage of female students gender, race, attendance status, student high school GPA, student tuition and fees hav odds of degree or certificate completion. Similar to the previous analysis in Table 4 6, institution size has a signifi degree or certificate completion. Holding other variables constant, students who attend a lager institution are 43.5% less likely to graduate than their counterparts in relatively smaller two year insti tutions. The percentage of female students is negatively related to s degree or certificate completion. A unit increase in percentage of female students leads to 5.35% decrease in student likelihood of graduation. C ompared with the analytical results presented in Table 4 6, more independent variables pertain ing to student college
75 experiences and student characteristics have positive impact on student academic achievement. Female students are in a better position in t erms of degree or certificate completion and they are 46.4% more likely to graduate than their male counterparts when studying the data between the year 20 03/04 and the year 20 08/09 In addition, minority students are 42.8% less likely to achieve student d egree or certificate completion than their W hite peers. Similar to the multilevel logistic regression analysis summarized in Table 4 likelihood of degree or certificate completion. A unit incr ease in student high school GPA result s when examining the data between the year 20 03/04 and the year 20 08/09 Students who attend community college s full time are 70.8% more likely to graduate than thei r peers who are not full time students when investigating the data between the year 20 03/04 and the year 20 08/09 Student tuition and fees ha ve a small of associate degree or certificate attainment in two year commu nity colleges. A unit completion between the year 20 03/04 and the year 20 08/09 Table 4 7. Final estimation of fixed effects Fixed Effect Coefficient Standard error t ratio Approx. d.f. p value For INTRCPT1, 0 INTRCPT2, 00 0.485 1.215 0.399 4 0 0.692 PROPPTFA, 01 0.0004 0.002 0.154 4 0 0.878 INSTITUT, 02 0.435 0.182 2.393 4 0 0.021* URBAN, 03 0.066 0.140 0.475 4 0 0.637 PCTMINOR, 04 0.0003 0.003 0.074 4 0 0.942 PCTWOMEN, 05 0.005 0.016 3.295 4 0 0.002* For INTRCPT1, 0 TUITIONF, 06 0.0001 0.0001 1.114 4 0 0.272 PELLGRAN, 07 0.00002 0.0002 0.145 4 0 0.886
76 Table 4 7. Continued Fixed Effect Coefficient Standard error t ratio Approx. d.f. p va lue PROPORTI, 08 0.010 0.006 1.732 4 0 0.091 INSTRUCT, 09 0.00001 0.00002 0.528 4 0 0.600 For FEMALE slope, 1 INTRCPT2, 10 0.464 0.106 4.375 187 0 0.000* For MINORITY slope, 2 INTRCPT2, 20 0.428 0.130 3.290 187 0 0.001* For TOTAI D slope, 3 INTRCPT2, 30 0.000002 0.00002 0.079 187 0 0.937 For JOBHOUR2 slope, 4 INTRCPT2, 40 0.002 0.003 0.500 1870 0.617 For FULLTIME slope, 5 INTRCPT2, 50 0.708 0.123 5.777 1870 0.000* For MEET slope, 6 INTRCPT2 60 0.300 0.260 1.150 187 0 0.250 For TALK slope, 7 INTRCPT2, 70 0.118 0.174 0.680 187 0 0.496 For HCGPAREP slope, 8 INTRCPT2, 80 0.138 0.052 2.667 40 0.011* For TEACTDER slope, 9 INTRCPT2, 90 0.066 0.124 0.535 10 0.606 For TESATDER slope, 10 INTRCPT2, 100 0.001 0.003 0.465 10 0.655 For TUITION slope, 11 INTRCPT2, 110 0.0001 0.00006 2.406 5 20 0.016* P values (p<.05) with indicate variables that are significant predictors for the outcome variab multiple imputed datasets. Variables pertain ing to student financial loans, average expenditure for academic support per FTE, average expenditure for student services per FTE and ave rage expenditure for institutional support per FET are not selected because of extremely low response rate. Chapter Summary This chapter presents the analytical results of the constructed multilevel logistic regression models presented in the methodolog y section. The models without any predictors (unconditional models) are firstly run to determine whether there is a need for running the multilevel logistic regression models. Given the variances across community colleges are significant, conditional model s are performed to account for institutional
77 level variances The analytical results indicate institutional level variables such as the percentage of part time faculty, the physical location of community colleges, and institution size ha ve significant imp completion between the year 20 03/04 and the year 20 05/06 And individual level variables such as working hours, high school GPA, full time attendance and tuition and fees are significant predictors for when examining the data between the year 2003/04 and the year 2005/06 between the year 20 03/04 and the year 20 08/09 institutional variables such as the percentage of female students and institution size are sig degree or certificate completion. In addition, student level variables such as gender, race, high school GPA, full time attendance, and t uition and fees are s ignificant predictors for student of degree or certificate completion. Although the variances are significantly reduced, i t should be noted that chi squares for random effects are still significa nt even after including studen t and institutional level variables in the multilevel logistic models ( =60.84, p<.05; =64.25, p<.05). This indicates that more student and/or institutional level variables should be included in the model s to reduce the variances in the dependent variables Table 4 8 : F inal estimation of variance components for pratt3yij (conditional) Random Effect Standard Deviation Variance Component d.f. 2 p value INTRCPT1, u 0 0. 20 0. 0 3 9 4 0 60.84 0.023 *The chi square multiple imputed datasets.
78 Table 4 9 : F inal estimation of variance components for pratt6yij (conditional) Random Effec t Standard Deviation Variance Component d.f. 2 p value INTRCPT1, u 0 0. 17 0. 03 4 0 64.25 0.012 *The chi square estimates are calculated in HLM 7.0 work on multiple imputed datasets.
79 CHAPTER 5 CONCLUSIONS AND IMPLICATIO NS In this section, analytical results from previous chapter are summarized to address three proposed research questions. Variables pertain ing to institutio nal and certifi cate completion are briefly reviewed and explanations for their significance are provided. Exemplary programs and activities that are eff ective in student retention and degree or certificate completion are enumerated respectively. In addition, implications for higher education research and administration are succinctly summarized. Institutional Characteristics As indicated by the results, employing a higher percentage of part time faculty has no negative impact upon academic achievement in two year community colleges. Contrary to the previous studies conducted by Benjamin (2002) and Jacoby (2006), the analytical result s from multilevel logistic regression analysis suggest that part time faculty have no negative impact upon student degree or certificate attainment in two year community colleges One possi ble explanation for the non negative impact of part tim completion is that community colleges hire a significant percentage of part time faculty who come direct ly from professional fields and have practical exp eriences, skills, and knowledge (Armstrong, 2005), which may help students achieve degree or certificate completion in two year community colleges. In addition, part time fac ulty ma y provide students connections to workplace or a community (McGuire, 1993 as cited in Armstrong, 2005).
80 Consistent with prior retention models proposed by Bean (1980), Swail, Redd, and Perna (2003), and Terenzini and Reason (2005), the current study reve als that other institutional characteristics also certificate completion. In particular, the percentage of female students, the physical chance of degree or certificate completion. Specifically, t he percentage of female students in a community college has when examining the data between the year 20 03/04 and the year 20 05/06 This is in contradiction with the prior research studies conducted by Pascarella and Terenzini (1991) Porter (2000), Scott, Bailey, and Kienzl (2004). However, this research result is in line with research study conducted by Bai ley and his collea gues (2005). One possible explanation is that female s tudents are more likely to dropout community colleges because of family commitments. Many student mothers reported that child care responsibility was the main reason why they leave community colleges wi thout completing associate degree or certificate (Rose & Hill, 2013) In addition, about 61% female students are part time students, a characteristic that is associated with low level of degree or certificate completion rate ( Bailey, Calcagno, Jenkins, Kie nzl, & Leinbach, 2005 ; Tinto, 2006). In addition, the current study also reveals the geographical location of a community college has a significant impact on student likelihood of degree or certificate completion when analyzing the data between the year 03/04 and the year 20 05/06 Students who pursue their studies in community colleges located in town suburban or city areas have a higher likelihood of degree or certificate completion than
81 their peer students who study in rural community colleges This i s different from research conducted by Calcagno and his colleagues (2008), in w hich the researchers claimed the geographical location did not have a significant impact on student degree or certificate completion One possible explanation is that community colleges located in rural areas are less likely to have sufficient res ources to ensure student success which result from low tax revenue and state funding and limited access to qualified staff ( Hicks & Jones, 2011 ) Institution size is inversely related to student degree or certificate completion when examining the data between the year 20 03/04 and the year 20 05/06, as well as the data between the year 20 03/04 and the year 20 08/09. Pursing academic studies in smaller institutions increases s rtificate completion. This is in line with the research conducted by Bailey and his colleagues, and Pascarella and Terenzini as smaller institutions might foster socially and academically integrated environment ( Bailey, Calc agno, Jenkins, Kienzl, & Leinbach, 2005; Pascarella & Terenzini, 1991, 2005 ). Other institutional characteristics such as the percentage of part time students, the percentage of minority students and average expenditure for instruction have no impact on s probability of degree or certificate completion. Individual Level Variables Student Demographics and Prior Experiences Consistent with student retention models proposed by Bean (1980) and Terenzini and Reason (2005), the current study reveals th at student demographics and prior experiences play an indispensable role in student or certificate completion in two year community colleges. Specifically, high school GPA is
82 degree or c ertificate completion. This concurs with the research conducted by Bettinger and Long (2007 as cited in Dickert Colin & Rubenstein, 2007). Students who entered t he community college with high er high school GPA are more lik ely to graduate than their counter parts who have low er high school GPA. One way to improve student degree or certificate completion is to set high academic admission standards for incoming students which is the common practice for four year institutions Yet, such policy is in direct conf lict with open access mission of community colleges and it is not an option for community colleges (Bailey, Calcagno, Jenkins, Kienzl, & Leinbach, 2005). One feasible option for two year community colleges is to offer effective development education prog rams ( i.e. math, reading, writing) to help unprepared students successfully adjust themselves to the college academic environmen t. While students can enter community colleges without proper academic requirements they should have a certain level of academi c preparation when purs u ing a certain degree or certificate. Another effective approach is to develop learning communities in com munity colleges, which is supported by prior research studies (Bailey & Alfonso, 2005 ; Wild & Ebbers, 2002 ). The learning commu nities are characterized as cohort model and the instruction is often centered on themes, issues, or pers pectives (Hagedorn, 2010). In learning communities, s tudents provide mutual support during their academic studies and extend their studies beyond class rooms (Hagedorn, 2010). These learning communities provide the structure for students to be integrated and engaged in learning (Wild & Ebbers, 2002). This is helpful given a significant portion of students are part
83 time commuter students who have little op portunity to interact with faculty members after class. In addition minority students are less likely than their W hite counterparts to complete associate degree or certificate in two year community colleges. Indeed, the achievement gap between minority s tudents and their White peers is of great concern in research literature (Bowen, Kurzweil, Tobin, 2005; Swail, Redd, & Perna, 2003). Congruent with prior research stu dies, the current research indicates male students are disadvantaged in terms of associa te degree or certificate completion Male students are less likely than their female co unterparts to achieve degree or certificate completion when analyzing the data between the year 20 03/04 and the year 20 08/09 Educational programs and policies should be implemented to help male students achieve degree or certificate completion. Student College Experience s Consistent with student retention models proposed by Bean (1980) and Terenzini and Reason (2005) variables pertain ing to student college experience s are found to influence student degree or certificate completion in two year community colleges. In particular, student working hours per week, tuition and fees, and student attendance status are found to have significant impact s on student ds of degree or certificate completion In community colleges, a considerable number of students are part time students, and their chance of degree or certificate completion is negatively impacte d by their attendance pattern. Community colleges should desi gn effective tim e attendance. Institutional aid should be implemented to help students make the transition from part time attendance to full time attendance.
84 mic studies exerts a colleges may create some on campus job opportunities to allow students to make some money and devote more time in their academic studies. In addition, stud ent tuition and fees have a s likelihood of degree or certificate completion when examining the data between the year 20 03/04 and the year 20 05/06, and the data between the year 20 03/04 and the year 20 08/09. Students might be motivated to achieve degree or certificate completion early with the increase in college tuition and fees. However, student college experience such as faculty student interaction is found to have no impact on student degree or certificate completion. T his is inconsistent with the model proposed by Tinto (1975), which proposes that student faculty interaction was a core component of student academic integration. One explanation is that most students in community colleges have limited interaction with the ir faculty members after class and faculty student interaction plays a limited role in promoting student of degree or certificate completion in two year community colleges. Comparing the Analytical Results between Three Year Data and Six Year s Data likelihood of degree or certificate completion between the year 20 03/04 and the year 20 05/06 are likelihood of degree or certificate comp letion between the year 20 03/04 and the year 20 08/09 The percentage of part time faculty in one community college has a small positive impact on student degree or certificate completion between the year 20 03/04 and the year 20 05/06. Y et, it has no impact on degree or certificate completion between the year 20 03/04 and the
85 year 20 08/09. Similarly, pursing academic studies in a non rural community college has a positive impact on student probability of degree or certificate completion whe n examining the data between the year 2003/04 and the year 2005/06. Yet, it has no odds of degre e or certificate completion when examining the data between the year 2003/04 and the year 2005/06. certificate completion between the year 2003/04 and the year 2008/09. On the other hand, being a femal e student in a community college has no significant impact on student degree or certificate completion when examining the data between the year 20 03/04 and the year 20 05/06. Yet, it has a small positive impact on student degree or certificate com pletion between the year 20 03/04 and the year 20 08/09 of degree or certificate completion when investigating the data between the year 20 03/04 and the year 2005/06. Yet being a minority student has a negative impact on in community colleges when studying the data between the year 20 03/04 and the year 20 08/09. The percentage of female students in a community college has no impac or certificate completion. Variables such as high school GPA, institution size, full time attendance, and tuition and fees consist completion in both datasets.
86 Table 5 1 : P degree or certificate completion Implications Implications for Higher Education Research Although the current study i s able to investigate whether the overall percentage of part degree or certificate completion it is time faculty during their academic studies. Research indicates the individual exposure to part time faculty rather than the overall percentage of part time faculty may play a bigger role in student retention and success (Burgess & Samuels, 1999; Harrington & Schibik, 2001; Jaeger & Eagan, 2009). In these studies, the instruction quality of part time faculty and that of full time faculty was compared. Burgess and Samuels (1999) found college students who took the first course from the part time instructor and took a second one from full time instructor seemed unprepared and were less likely to earn good grade in the second class. Yet, more research studies are needed to validate this conclusion as these research studies were conducted at one institution of higher education Data pertain ing t ime faculty are not available for the community colleges under study. One potential direction for the research is to collect information associated time faculty on these community colleges. Such information would provide imp ortant information regarding the effect of part time faculty on studen t of degree or certificate completion.
87 Secondly f uture research should consider combing National Study of Postsecondary Faculty dataset ( NSOPF ), Integrated Postsecondary Education Data System ( IPEDS), and Beginning Postsecondary Study Longitudinal Study ( BPS). T he NSOPF dataset includes faculty characteristics which can be included into multi level logistic regression analysis to account for variances in dependent variabl e s and yield more accurate parameter estimates. Research studies on student persistence and graduation primarily center on traditional aged students in four year institutions. Wh ile the se proposed student retention models shed light on retention studies, educational research should be furthered to investigate how to ensure student degree or certificate completion in two year institutions of higher education. Student retention models pertain ing to community college students should be proposed base d on the c urrent study. In addition, qualitative research should be conducted to study whether part time faculty has an impact upon student learning in community colleges. Such research is critical for researchers and practitioners when evaluating the overall impac t of part time faculty on student learning outcomes (i.e. persistence, degree or certificate completion) Implications for Higher Education Administration their degree or ce rtificate completion in community colleges. Community college leaders and administrators should work collaboratively to identify academically unprepared students as research studies suggested that identifying academically at risk students early will effect ively improve student retention and graduation (Beck & Davidson, 2001; Reisberg, 1999 as cited in College Retention Report, 2011). In addition to placement tests, community college leaders should establish early warning programs for the first
88 year students administrative offices should be established to identify these students. And education interventions such as tutoring and mentoring services should be in place to help these students to ad just the college life in addition to developmental or remedial education programs ( Wild & Ebbers, 2002) time faculty have a negative impact on student degree or certificate c ompletion college leaders and administrators should still be cautious. Prior research studies confirmed that faculty satisfaction is linked with faculty performance and part time faculty have low level satisfaction due to unfavorable working conditions (J acoby, 2006 ; Weiss & Pankin, 2011 ). As such, community college administrator s and leaders should design effective policies that can improve working environment and comp ensation for part time faculty. As indicated by the current study, minority students are less likely to achieve degree or certificat e completion than their W hite peers. A positive racial climate on community college campus is important for minority students to persist and achieve degree or certificate completion (Cabrera, Nora, Terenzini, Pas carella, & Hagedorn, 1999; College Borad, 2010). Effective inventi on strategies should be used in classroom or college level (i.e. collaborative classroom learning practices, multi culture curriculum, and cultural awareness workshops) so that students, fac ulty, administrators, and staff can develop experiences and skills for a positive racial climate on campus and contribute to student and organizational learning(College Board, 2010). Concluding Thoughts Despite the study limitations, the current study co ntributes the current research on student retention and degree or certificate completion By matching students with
89 their two year community colleges, the current study fills the research gap about whether part time faculty have an impact upon student degr ee or certificate completion using student and institutional level data. In addition, the current research examines variables pertain ing to student background and academic preparation, college experiences, and institutional level characteristics. Effective programs and services can be implemented to address significant factors that strongly predict student degree or certificate completion. The most noticeable finding is that high school GPA is a strong predictor for student degree or certificate completion in two year c ommunity colleges. Students who are academically prepared are more likely to obtain associate degree or certificate than their ill prepared counterparts The physical location of a community de gree or certificate completion Community colleges in rural areas are less like ly to get sufficient funding as compared to their counterparts in town, cities, or suburban areas. More resources should be devote d to retain ing students in rural community coll eges. For many of community college students, attending community college is their only way to attend post secondary educat ion institutions ( Calcagno, Bailey, Jenkins, Kienzl, & Leinbach, 2008; Cohen & Brawer, 2003 as cited in Schuetz, 2005; Hagedorn, 2010). Educational researchers and practitioners should work collaboratively to promote student degree or certificate completio n by address ing barriers pertain ing to the above mentioned variables
90 APPENDIX A Criteria for Se lecting Community Colleges Miscellaneous Indicators U.S. only; Title IV participating State or other jurisdiction Alabama; Alaska; Arizona; Arkansas; Califor nia; Colorado; Connecticut; Delaware; District of Columbia; Florida; Georgia; Hawaii; Idaho; Illinois; Indiana; Iowa; Kansas; Kentucky; Louisiana; Maine; Maryland; Massachusetts; Michigan; Minnesota; Mississippi; Missouri; Montana; Nebraska; Nevada; New Ha mpshire; New Jersey; New Mexico; New York; North Carolina; North Dakota; Ohio; Oklahoma; Oregon; Pennsylvania; Rhode Island; South Carolina; South Dakota; Tennessee; Texas; Utah; Vermont; Virginia; Washington; West Virginia; Wisconsin; Wyoming; American Sa moa; Federated States of Micronesia; Guam; Marshall Islands; Northern Marianas; Palau; Puerto Rico; Virgin Islands Geographic region US Service schools; New England CT ME MA NH RI VT; Mid East DE DC MD NJ NY PA; Great Lakes IL IN MI OH WI; Plains IA KS MN MO NE ND SD; Southeast AL AR FL GA KY LA MS NC SC TN VA WV; Southwest AZ NM OK TX; Rocky Mountains CO ID MT UT WY; Far West AK CA HI NV OR WA; Outlying areas AS FM GU MH MP PR PW VI Sector Public, 2 year; Private not for profit, 2 year; Private for profit, 2 year Degree granting status Degree granting Highest degree offered Associates; Associates and first professional Carnegie Classification 2000 Associates Colleges Degree of urbanization (Locale)
91 Large city; Mid size city; Urban fringe of large city; Urba n fringe of mid size city; Large town; Small town; Rural; Not assigned Has full time first time undergraduates Yes
92 APPENDIX B R Code for Combing D atasets and Normalizing Weights library(mice) library(lme4) #Merging the data with the unit ID setwd("G:/Par ttimeFaculty/datasets/original data") data.bps=read.csv(file="bps.csv") data.ipeds=read.csv(file="ipeds.csv") ipedsyu=read.csv(file="ipedsyu.csv") ptfp=read.csv(file="ptfp.csv") ipeds=merge(ipedsyu,ptfp,by="NPSIPDID") write.csv(ipeds,file="ipeds.csv") save (ipeds, file="ipeds.Rdata") dim(ipeds) names(ipeds) summary(ipeds) head(ipeds) bps=read.csv(file="bps.csv") bpsipeds=merge(bps,ipeds,by="NPSIPDID") write.csv(bpsipeds,file="bpsipeds.csv") save(bpsipeds,file="bpsipeds.Rdata") #Summarizing the data dim(bps ipeds) names(bpsipeds) summary(bpsipeds) head(bpsipeds) #Identifying the missing values in variables bpsipeds=read.csv("bpsipeds.csv",na.strings=c( 1: 9, "NA")) na.strings=c( 1: 9,"NA") missing.indicator = is.na(bpsipeds) head(missing.indicator) missing. summary.variables = apply(missing.indicator,2,sum) missing.summary.cases = apply(missing.indicator,1,sum) table(missing.summary.variables) table(missing.summary.cases) #Separating the variables with missing values from those without missing values missin g.code = missing.summary.variables == 0 data.not.missing = bpsipeds[,missing.code] dim(data.not.missing) names(data.not.missing)
93 data.with.NA = bpsipeds[,!(missing.code)] names(data.with.NA) #Combining the dataset with missing values with the dataset wit hout missing values data= cbind(data.with.NA,data.not.missing) #Converting numeric to factor if the variables are categorical data data$PRATT3Y=factor(data$PRATT3Y) data$PRATT6Y=factor(data$PRATT6Y) data$GENDER=factor(data$GENDER) data$RACECEN=factor(data $RACECEN) data$HCGPAREP=factor(data$HCGPAREP) data$ATTEND=factor(data$ATTEND) data$institutionsize=factor(data$institutionsize) data$urban=factor(data$urban) data$FREQ04A=factor(data$FREQ04A) data$FREQ04B=factor(data$FREQ04B) class(data$PRATT3Y) class(dat a$PRATT6Y) class(data$GENDER) class(data$RACECEN) class(data$HCGPAREP) class(data$ATTEND) class(data$institutionsize) class(data$urban) class(data$FREQ04A) class(data$FREQ04B) # Conducting multiple imputation and save the multiple imputed datasets imputed .partial.data = mice(data[,1:10],m = 5,diagnostics = F) imputed.partial.data1 = complete(imputed.partial.data,1) imputed.partial.data2 = complete(imputed.partial.data,2) imputed.partial.data3 = complete(imputed.partial.data,3) imputed.partial.data4 = compl ete(imputed.partial.data,4) imputed.partial.data5 = complete(imputed.partial.data,5) write.csv(imputed.partial.data1,"imputed.partial.data1.csv") write.csv(imputed.partial.data2,"imputed.partial.data2.csv") write.csv(imputed.partial.data3,"imputed.partial .data3.csv") write.csv(imputed.partial.data4,"imputed.partial.data4.csv") write.csv(imputed.partial.data5,"imputed.partial.data5.csv")
94 data1=cbind(data[11:34],imputed.partial.data1) data2=cbind(data[11:34],imputed.partial.data2) data3=cbind(data[11:34],i mputed.partial.data3) data4=cbind(data[11:34],imputed.partial.data4) data5=cbind(data[11:34],imputed.partial.data5) setwd("G:/ParttimeFaculty/datasets/original data") write.csv(data1, file="data1.csv") write.csv(data2, file="data2.csv") write.csv(data3, file="data3.csv") write.csv(data4, file="data4.csv") write.csv(data5, file="data5.csv") # Recoding the values in dependent variables and normalizing the weight data1=read.csv(file="data1.csv") data1$PRATT3Y[data1$PRATT3Y==1]=1 data1$PRATT3Y[data1$PRATT3Y ==2]=1 data1$PRATT3Y[data1$PRATT3Y==3]=0 data1$PRATT3Y[data1$PRATT3Y==4]=0 table(data1$PRATT3Y) data1$PRATT6Y[data1$PRATT6Y==1]=1 data1$PRATT6Y[data1$PRATT6Y==2]=1 data1$PRATT6Y[data1$PRATT6Y==3]=0 data1$PRATT6Y[data1$PRATT6Y==4]=0 table(data1$PRATT6Y) da ta1$norm.w = data1$WTB000/mean(data1$WTB000) write.csv(data1, file="data1.csv") data2=read.csv(file="data2.csv") data2$PRATT3Y[data2$PRATT3Y==1]=1 data2$PRATT3Y[data2$PRATT3Y==2]=1 data2$PRATT3Y[data2$PRATT3Y==3]=0 data2$PRATT3Y[data2$PRATT3Y==4]=0 table( data2$PRATT3Y) data2$PRATT6Y[data2$PRATT6Y==1]=1 data2$PRATT6Y[data2$PRATT6Y==2]=1 data2$PRATT6Y[data2$PRATT6Y==3]=0 data2$PRATT6Y[data2$PRATT6Y==4]=0 table(data2$PRATT6Y) data2$norm.w = data2$WTB000/mean(data2$WTB000) write.csv(data2, file="data2.csv")
95 data3=read.csv(file="data3.csv") data3$PRATT3Y[data3$PRATT3Y==1]=1 data3$PRATT3Y[data3$PRATT3Y==2]=1 data3$PRATT3Y[data3$PRATT3Y==3]=0 data3$PRATT3Y[data3$PRATT3Y==4]=0 table(data3$PRATT3Y) data3$PRATT6Y[data3$PRATT6Y==1]=1 data3$PRATT6Y[data3$PRATT6Y==2 ]=1 data3$PRATT6Y[data3$PRATT6Y==3]=0 data3$PRATT6Y[data3$PRATT6Y==4]=0 table(data3$PRATT6Y) data3$norm.w = data3$WTB000/mean(data3$WTB000) write.csv(data3, file="data3.csv") data4=read.csv(file="data4.csv") data4$PRATT3Y[data4$PRATT3Y==1]=1 data4$PRATT3Y [data4$PRATT3Y==2]=1 data4$PRATT3Y[data4$PRATT3Y==3]=0 data4$PRATT3Y[data4$PRATT3Y==4]=0 table(data4$PRATT3Y) data4$PRATT6Y[data4$PRATT6Y==1]=1 data4$PRATT6Y[data4$PRATT6Y==2]=1 data4$PRATT6Y[data4$PRATT6Y==3]=0 data4$PRATT6Y[data4$PRATT6Y==4]=0 table(dat a4$PRATT6Y) data4$norm.w = data4$WTB000/mean(data4$WTB000) write.csv(data4, file="data4.csv") data5=read.csv(file="data5.csv") data5$PRATT3Y[data5$PRATT3Y==1]=1 data5$PRATT3Y[data5$PRATT3Y==2]=1 data5$PRATT3Y[data5$PRATT3Y==3]=0 data5$PRATT3Y[data5$PRATT 3Y==4]=0 table(data5$PRATT3Y) data5$PRATT6Y[data5$PRATT6Y==1]=1 data5$PRATT6Y[data5$PRATT6Y==2]=1 data5$PRATT6Y[data5$PRATT6Y==3]=0 data5$PRATT6Y[data5$PRATT6Y==4]=0 table(data5$PRATT6Y) data5$norm.w = data5$WTB000/mean (data5$WTB000) write.csv(data5, fil e="data5.csv")
96 REFERENCE LIST ACT.(2013).2012 Retention/Completion Summary Tables. Retrieved from http://www.act.org/research/policymakers/pdf/12retain_trends.pdf Alexand er, A., Karvonen, M., Ulrich, J. Davis, T., & Wade, A. (2012). Community college faculty competencies. Community College Journal of Research & Practice, 36 (11), 849 862. doi:10.1080/10668926.2010.515511 Alon, S. (2005). Model mis specification in assessin g the impact of financial aid on academic outcomes. Research in Higher Education 46(1), 109 125. American Association of Community Colleges. (2010). American association of community colleges 2010 fact sheet American Association of Community Colleges, Wa shington, DC 20015. American Association of Community Colleges. (201 2 ). American association of community colleges 2010 fact sheet American Association of Community Colleges, Washington, DC 20015. American Association of Community Colleges. (2013 ). Ameri can association of community colleges 2010 fact sheet American Association of Community Colleges, Washington, DC 20015. American Federation of Teachers. (2011). Student success in higher education. Retrieved from http://www.aft.org/pdfs/highered/studentsuccess0311.pdf Armstrong, D. (2005). Part time Faculty in the Florida Community College System: A Status Report. Florida Department of Education. Retrieved from http://www.fldoe.org/fcs/Vision/PDFs/PR2005_2.pdf Astin, A. W. (1975). Preventing Students from Dropping Out. San Francisco: Jossey Bass. Astin, A. W. (1999). Student involvement: A developmental theory for hig her education. Journal of College Student Development, 40 (5), 518 529. Aud, S et al. (2011 ). The condition of education 2011 Washington, D.C.: U.S. Dept. of Education, National Center for Education Statistics. Retrieved from http://nces.ed.gov/pubs2011/2011033.pdf Aud, S et al. (2012 ). The condition of education 2012 Washington, D.C.: U.S. Dept. of Education, National Center for Education Statistics. Retrieved from http://nces.ed.gov/pubs2012/2012045.pdf
97 Bailey., T. R., & Alfonso, M. (2005). Paths to persistence: An analysis of research on program effectiveness at community colleges Indianapolis, IN: Lumina Foundation for Education. Bailey, T. Alfonso, M., Calcagno, J. C, Jenkins, D., Kienzl, G., & Leinbach, T. (2004). Improving student attainment in community colleges: Institutional characteristics and policies Community College Research Center, Teachers College, Columbia University, New Yor k, NY 10027. Bailey, T., Calcagno, J C Jenkins, D Kienzl, G & Leinbach, T (2005). The effects of institutional factors on the success of community college students. Community College Research Center, Teachers College, Columbia University, New York NY 10027. Bailey, T. R., & Aver ianova, I. E. (1998). Multiple missions of community colleges: Conflicting or complementary? Com munity College Research Center Teachers College, New York, NY 10027 Bailey,T.R., & Leinbach, D.T. (200 5). Is student success c onsidered institutional failure? The accountability debate at community colleges New York, NY: Com munity College Research Center, Teachers College, Columbia University. Bailey,T. R., Morest,V. S. (2004). The organizational efficiency of multiple missions f or community colleges Community College Research Center, Teachers College, Columbia University, New York, NY 10027. Barrett, I. K. (2012). Student and institutional variables that affect time to degree completion of community college students. ( Doctoral d issertation ). Retrieved from ProQuest Dissertations and Theses (UMI No. 3528731 ) Baum, S., & M a, J. (2012 ). Trends in college pricing 2012 New York: College Board. Baum, S., Ma, J. & Payea, K. (2010). Education pays 2010 the benefits of higher educatio n for individuals and society. Retrieved from http://advocacy.collegeboard.org/sites/default/files/Education_Pays_2010.pdf Beach, J. M. (2011). Gateway to opportu nity? A history of the community college in the United States. Sterling, VA: Stylus Publishing Bean, J. P. (1980). Dropouts and turnover: The synthesis and test of a causal model of student attrition. Research in Higher Education, 12 (2), 155 87. Beck, H. P., & Davidson W. D. (2001). Establishing an early warning system: P redicting low grades in college students from survey o f academic orientations scores. Research in Higher Education 42(6), 709 723.
98 Benjamin, E. (2002). How over reliance on contingent a ppointments diminishes faculty involvement in student learning. Peer Review 5(1), 4 10. Benoit, R. J., & Smith, A. (1980). Demographic and job satisfaction characteristics of Florida community college faculty. Community/Junior College Research Quarterly, 4 (3), 263 76. Berger, J B., & Braxton,J M. (1998). Revising Tinto's interactionalist theory of student departure through theory elaboration: Examining the role of organizational attributes in the persistence process. Research in Higher Education, 39 (2) 103 119. Bettinger, E. (2004). How financial aid affects persistence. In College choices: The economics of where to go, when to go, and how to pay for it (pp. 207 238) University of Chicago Press. Bettinger, E.P., & Long, B. (2006). The increasing use o f Adjunct Instructors at public institutions: Are we hurting students?. In R.Ehrenberg (Ed.), Assessing Public Higher Education at the Start of the 21st Century Bettinger, E. P., & Long, B.T. (2010). Does cheaper mean better? The impact of using adjunct in structors on student outcomes. The Review of Economics and Statistics, 92 (3), 598 613. Birnbaum, R. (1988). How colleges work: The cybernetics of academic organization and leadership San Francisco: Jossey Bass. Boggs, G. R. (2012). The evolution of the community college in America: Democracy's colleges. Community College Journal, 82 (4), 36 39. Bowen, W G., Chingos, M M. & McPherson, M S. (2009). Crossing the finish line: Completing College Princeton, N.J.: Princeton University Press. Bowen, W G., Kurzweil, M A., Tobin, E M., & Pichler, S C ( 2005). Equity and excellence in American higher education Charlottesville: University of Virginia Press. Boyer, E L ( 1987). College: The undergraduate experience in America New York: Harper & Row. Braxton, J M., Hirschy, A S., & McClendon S A. (2004). Understanding and reducing college student departure San Francisco.: Jossey Bass.
99 Brint, S. G., & Karabel, J. (1989). The diverted dream: Community colleges and the promise of educational opportunity in America, 1900 1985 New York: Oxford U niversity Press. Brunn, R. J. (2009). Intersectionality at work: Race, class, and gender gaps in post secondary achievement and attainment. ( Doctoral dissertation ). Retrieved from ProQuest Dissertations and Theses databases (UMI No. 3363260 ) Buchmann, C & DiPrete, T A ( 2006). The growing female advantage in college completion: The role of family background and degree or certificate completion American Sociological Review. 71 (4), 515. Burdman, P. (2005). The student debt dilemma: Debt aversion as a bar rier to college access. Retrieved from: http://cshe.berkeley.edu/publications/publications.php?s=1 Burgess, L. A. (1999). Impact of full time versus part time instructor status on college student retention and academic performance in sequential courses. Co mmunity College Journal of Research and Practice, 23 (5), 487 498. doi:10.1080/106689299264684 Burke, W. W., & Litwin, G. H. (1992). A causal model of organizational performance and change. Journal of Management, 18 (3), 523 545. doi:10.1177/014920639201800 306 Byun, S Irvin, M J., & Meece,J L.(2012). Predictors of bachelor's degree completion among rural students at four year institutions. Review of Higher Education, 35 (3), 463 484. Cabrera, A. F., Nora, A., & Castaeda, M.B. (1993). College persistence : Structural equations modeling test of an integrated model of student retention. Journal of Higher Education, 64 (2) 123. Calcagno, J. C., Bailey, T., Jenkins, D., Kienzl, G., & Leinbach, T. (2008). Community college student success: What institutional ch aracteristics make a difference? Economics of Education Review, 27 (6), 632 645. doi:10.1016/j.econedurev.2007.07.003 Calcagno, J. C., Crosta, P., Bailey, T., & Jenkins, D. (2007). Stepping stones to a degree: the impact of enrollment pathways and mileston es on community college student outcomes. Research in Higher Education 48( 7 ) 775 801. Carroll, B., & Blumen, M. W. (1973). Job satisfaction; a review of the literature. Ithaca: New York State School of Industrial and Labor Relations, Cornell University.
100 Carnegie Foundation for the Advancement of Teaching. (2005) The Carnegie Classification of Institutions of Higher Education Retrieved from http://classifications.carnegiefoundation.org/ Ca taldi, E. F. Fahimi, M., Bradburn, E. M. & Zimbler, L. (2005). 2004 national study of postsecondary faculty (NSOPF: 04) report on faculty and instructional staff in fall 2003. Retrieved from http://nces.ed.gov/pubsearch/pubsinfo.asp?pubid=2005172 Center for Community College Student Engagement. (2012). A matter of degrees promising practices for community college student success: A first look. Retrieved from http://www.ccsse.org/docs/MatterofDegrees.pdf Chambel, M., & Curral, L. (2005). Stress in academic life: Work characteristics as predictors of student well being and performance. Human Resources Abstracts, 40 (2) Chang, J. C. (2005). Faculty student interaction at the community college: A focus on students of color. Research in Higher Education, 46 (7), 769 802. doi:10.1007/s11162 004 6225 7 Charlier, H. &Williams, M. (2011). The reliance on and demand for adjunct faculty m embers in america's rural, suburban, and urban community colleges. Community College Review, 39 (2), 160 180. Cheng, D. X., & Alcantara, L. (2004). To work or not to work: The impact of work on students' college experience. Retrieved from http://www.eric.ed.gov/ERICWebPortal/search/detailmini.jsp?_nfpb=true&_&ERI CExtSearch_SearchValue_0=ED49 0999&ERICExtSearch_SearchType_0=no&a ccno=ED490999 Cohen, A. M. (1998). The shaping of American higher education: Emergence and growth of the contemporary system San Francisco: Jossey Bass. Cohen, A. M.(2001). Governmental policies affecting community c olleges: a historical perspective. In Townsend, B. & Twombly, S.(Eds.), Community colleges : policy in the future context (pp.1 22). Westport, CT: Ablex Publishing. Cohen, A. M., & Brawer, F. B. (2003). The American community college San Francisco: Josse y Bass. Cohen, A. M., & Brawer, F. B. (2008). The American community college San Francisco: Jossey Bass. Coley, Richard J (2000). The American community college turns 100: A look at its students, programs, and prospects Princeton, N.J.: Educational Te sting Service.
101 College and University Professional Association for Human Resources (2001). 2000 2001 national faculty salary survey by discipline and rank in public four year colleges and universities Washington, D.C.: College and University Professional Association for Human Resources. College Board. (2010). Securing the Future. Retention Models in Community Colleges Retrieved from http://advoca cy.collegeboard.org/sites/default/files/community college security future retention models 5875.pdf College Board Advocacy & Policy Center (2011). How four year colleges and universities organize themselves to promote student persistence: The emerging na tional picture College Board Advocacy & Policy Center. NY 10023. Community College Survey of Student Engagement.(2012). 2012 Cohort Key Findings Retrieved from http://www.ccsse.o rg/survey/survey.cfm Curs, B R., & Harper, C E ( 2012). Financial aid and first year collegiate GPA: A regression discontinuity approach. Review of Higher Education, 35 (4), 627 649. Curtis, J.W., & Jocobe, M. (2006). Consequences: An Increasing Conting ent Faculty AAUP Contingent Faculty Index 2006 Retrieved from http://www.aaup.org/AAUP/pubsres/research/conind2006.htm DesJardins, S. L., Ahlburg, D.A., McCall,B. P. (2002). A temporal investigation of factors related to timely degree completion. The Journal of Higher Education, 73 (5), 555 581. D e Valero, Y. F. (2001). Departmental factors affecting time to degree and completion rates of doctoral students at one land grant res earch institution. Journal of Higher Education, 341 367. Dickert Conlin, S., R ubenstein, R. H. (2007). Economic inequality and higher education Access, persistence, and success New York: Russell Sage Foundation. Dougherty, K.J. (1994). The contradictory college: Conflicting origins, impacts, and futures of the community college A lbany, NY: State University of New York Press. Dowd, A. C. (2008). Dynamic Interactions and Intersubjectivity: Challenges to Causal Modeling in Studies of College Student Debt. Review of Educational Research 78 ( 2 ) 232 259. Dowd, A. C. & Coury, T. (2006). The effect of loans on the persistence and attainment of community college students. Research in Higher Education, 47 (1), 33 62.
102 Dundes, L. & M arx, J. (2006). Balancing work and academics in college: Why do students working 10 19 hours per week excel? Jo urnal of College Student Retention, 8 (1), 107 120. Dynarski, S. M. (2005). Finishing college: The role of schooling costs in degree attainment Working paper, Harvard University Eagan, M. K. (2008). Closing the gate: Part time faculty instruction in gat ekeeper courses and first year persistence. New Directions for Teaching & Learning, (115), 39 53. doi:10.1002/tl.324 Eagan, M. K & Jaeger, A. J. (2009). Effects of exposure to part time faculty on community college transfer. Research in Higher Education, 50 (2), 168 188. Education Commission of the States. (2003). Closing the College Participation Gap: A National Summary Denver: The Center for Community College Policy, Education Commission of the States. Ehrenberg, R G. & Zhang, L. ( 2005 ). Do Tenured a nd Tenure Track Faculty Matter? Journal of Human Resources 40(3) 647 659 Eurostat. ( 200 2). The Life of Women and Men in Europe: A St atistical Portrait. Luxembourg: Eurostat Fjortoft, N. F. (1995). College student employment: Opportunity or deterrent? P aper presented at the annual meeting of the American Educational Research Association, San Francisco, CA (ERIC Docum ent Reproduction Service No. ED 386079). Flores, S. M., & Park, T. J. (2013). Race, ethnicity, and college success: Examining the continued significance of the minority serving institution. Educational Researcher, 42 (3), 115 128. Forehand, G. A., & Von Haller, G. (1964). Environmental variation in studies of organizational behavior. Psychological Bulletin, 62 (6), 361 382. doi:10.1037/h0045960 French, W. L. & Bell, C. (1999). Organization development: Behavioral science interventions for organization improvement (6 th ed.). Upper Saddle River, NJ: Prentice Hall. Geller, H. A (2001). A brief history of community colleges and a personal view o f some issues (open admissions, occupational training and leadership). Retrieved from ERIC database. ( ED459881 )
103 Gladieux, L. E., & Perna, L. W (2005). Borrowers who drop out: A neglected aspect of college student loan trend San Jose : National Center for Public Policy and Higher Education. Glisson, C., & James, L. R. (2002). The cross level effects of culture and climate in human service teams. Journal of Organizational Behavior, 23 (6), 767 794. doi:10.1002/job.162 Goldin, C D., Katz, L. F., & Kuziemko, I (2006). The homecoming of American college women: The reversal of the college g ender gap Cambridge: National Bureau of Economic Research. Graham, J. W. (2012). Multiple imputation and analysis with multilevel (cluster) data. In m issing data (pp. 133 150) New York: Springer Grubb, N. W. (1999). Honored but Invisible: An Inside Look at Teaching in Community Colleges. New York: Routledge. Hagedorn, L. S. (2000). Conceptualizing faculty job satisfaction: Components, theories, and outcomes. New Directions for Institutional Research, (105), 5 20. doi:10.1002/ir.10501 Hagedorn, L. S. (2 010). The pursuit of student success: The directions and challenges facing community colleges. In Higher education: Handbook of theory and research (pp. 181 218). Springer Netherlands. Hagedorn, L. S., Chi, W. Cepeda, R. M., & McLain, M. (2006). An Invest igation of Critical Mass: The role of latino representation in the success of urban community college students. Research in Higher Education, 48 (1), 73 91. Hagedorn, L. S., Maxwell, W., Rodriguez, P., Hocevar, D., & Fillpot, J. (2000). Peer and student fa culty relations in community colleges. Community College Journal of Research & Practice 24(7), 587 598. Hanniford, B. E., & Sagaria, M A. (1994). The impact of work and family roles on associate and baccalaureate degree completion among students in early adu lthood. Retrieved from ERIC database. (ED370520). Harrington, C ., & Schibik, T. (2001). Caveat emptor: Is there a relationship between part time faculty utilization and student learning outcomes and retention? AIR 2001 annual forum paper. Retrieved from ERIC database. (ED456785). Haviland, D., Shin, S., & Turley, S. (2010). Now I'm ready: The impact of a professional development initiative on faculty concerns with program assessment. Innovative Higher Education, 35 (4), 261 275.
104 Hendrick, R. Z., Hightower, W. H., & Gregory, D. E. (2006). State funding limitations and community college open door policy: Conflicting priorities? Community College Journal of Research & Practice 30(8), 627 640. Herzberg, F., Mausner, B., & Snyderman, B. B. (1993). The motivation to work New Brunswick : Transaction Publishers. Hicks, C., & Jones, S. J. (2011). At Issue: Survival Tactics for Small, Rural Serving Community Colleges. Community College Enterprise, 17, 2, 28 45. Hofmann, D. A., & Gavin, M. B. (1998). Centering decision s in hierarchical linear models: Implications for research in organizations. Journal of Management 24(5), 623 641. Hurtado, S., & Carter, D. F. (1997). Effects of college transition and perceptions of the campus racial and social climate on Latino studen Sociology of Education ,70, 324 345. Hutton, J. B., & Jobe, M. E. (1985). Job satisfaction of community college faculty. Community Junior College Research Quarterly of Research and Practice, 9 (4), 317 324. doi:10.1080/0361697850090 402 Ingels, S. J et al (2013). Educat ion longitudinal study of 2002 ( 2002/12 ) Third Follow Up Field Test Report Retrieved from http://nces.ed.gov/pubsearch/pubsinfo.asp?pubid=201203 Integrated Postsecondary Education System. (2013). Glossary Retrieved from http://nces.ed.gov/ipeds/glossary/ Ishitani, T. T (2006). Studying attrition and degree completion behavior among first generati on college students in the United States. The Journal of Higher Education, 77 (5), 861 885. Jacobs, J. A. (1996). Gender inequality and higher education. Annual Review of Sociology, 22 153 185. Jacobs, J. A. (1999). Gender and the stratification of colle ges. The Journal of Higher Education, 70 (2), 161 187. Jacoby, D. (2006). Effects of part time faculty employment on community college graduation rates. The Journal of Higher Education, 77 (6), 1081 1103. Jaeger, A. J., & Eagan, M.K., (2009). Unintended conse quences: Examining the effect of part time faculty members on associate's degree completion. Community College Review, 36 (3), 167 194.
105 Johnson, I.Y. (2006). Examining part time faculty utilization and its impact on student retentions at a Public Research University Paper presented at the Annual Forum of the Association for Insti tutional Research, Chicago, IL. Judge T A ., Thoresen, C J, Bono J E, & Patton G K ( 2001). The job satisfaction job performance relationship: A qualitative and quantitative re view. Psychological Bulletin, 127 (3), 376 407. Kamens, Effects on occupational choice and college attrition. Sociology of Education 44 270 296. Khan, H., & Shaw, E. (2011). Multilevel logistic regr ession analysis applied to binary contraceptive prevalence data. Journal of Data Science, 9 93 110. Kim, M. M., & Conrad, C. F. (2006). The impact of historically black colleges and universities on th e academic success of African A merican students. Resear ch in Higher Education, 47 (4), 399 427. Kodrzycki, Y. K. (2004). College completion gaps between blacks and whites: What accounts for regional differences? New England Economic Review, 14 (2), 37 62. Kuh, G. D., Cruce, T. M., Shoup, R., Kinzie, J., & Gonye a, R. M. (2008). Unmasking the effects of student engagement on first year college grades and persistence. The Journal of Higher Education, 79 (5), 540 563. Lam, L. P. T. (1999). Assessing Financial Aid Impacts on Time to Degree for Nontransfer Undergradua te Students at a Large Urban Public University. AIR Annual Forum Paper. Paper presented at the 39 th Annual Forum of Association for Institutional Research, Seattle, WA. Leeuw, J. & Meijer, E. (2008). Handbook of multilevel analysis New York: Springer. L eslie, D. W., & Gappa J.M. ( 2002). Part time faculty: Competent and committed. New Directions for Community Colleges, 118 (118), 59 68. Levin, J. (2007). Multiple judgments: Institutional context and part time faculty. New Directions for Community Colleges 140, 15 20. Levin, J. S. (2004). The community college as a baccalaureate granting institution. Review of Higher Education, 28 (1), 1 22. Levin, J.S. (2012). Understandings of community colleges in need of resuscitation: The case of community college facult y, In John S. Levin & Susan Kater, S., (Eds.), Understanding Community Colleges New York: Routledge.
106 Levin, J. S., Kater, S., & Wagoner, R. L. (2006). Community college faculty: At work in the new economy New York: Palgrave Macmillan. Li, D. (2008 ). Degree Attainment of Undergraduate Student Borro wers in Four Year Institutions: A Multilevel Analysis. NASFAA Journal of Student Financial Aid 37(3), 5 16 Light, A. (2001). In School work experience and the returns to schooling. Journal of Labor Econom ics, 19 (1), 65 93. Linda, S. H. (2000). Peer and student faculty relations in community colleges. Community College Journal of Research and Practice, 24 (7), 587 598. doi:10.1080/10668920050139730 Little, R. & Rubin, D. B. (2002). Statistical analysis wi th missing data Hoboken, N.J.: Wiley. Locke, E. A. (1976). The nature and causes of job satisfaction In M.D. Dunnette (Ed.), Handbook of Industrial and Organizational Psychology 1297 1349. Long, B. T. (2008). What is known about the impact of financia l aid? Implications for policy. National Center for Postsecondary Research ( NCPR) Working Paper. Retrieved from http://www.postsecondaryresearch.org/i/a/document/6963_Long FinAid.pdf Long, B. T. (2010). Financial aid: A key to community college student success. The White House Summit on Community Colleges, 52. Lundy, V. C. (2010). The significance of interactions: Understanding gender, ethnicity/race, and socioeconomic st degree completion ( Doctoral dissertation) Retrieved from ProQuest Dissertations and Theses database ( UMI No. 3431147 ). McGuire, J. (1993). Part Time Faculty: Partners in Excellence. Leadership Abstracts 6( 6). McLaughlin, F. (2005). Adjunct faculty at the community colle ge: Second class professoriate? Teaching English in the Two Year College 33, 185 193. McPhail, C., & McPhail, I. (2006). Prioritizing community college missions: A directional effort. New D irections for Community Colleges, (136), 91 99. Mellow, G. O., & Heelan, C. (2008). Minding the Dream: The Process and Practice of the American Community College Lanham, MD: Rowman & Littlefield.
107 Merisotis, J. P & Phipps, R. A. (2000). Remedial education in colleges and universities: What's really going on? The Review of Higher Education, 24 (1), 67 85. Milano, J., Reed, B., & Weinstein, P. Jr., (2009 ). fastest growing jobs and why community college graduates will get them Washington, DC: The New Democratic Leadership Council. Retrieved from http://www.dlc.org/documents/DLC_HotJobs.pdf Mohammadi, J. (1994). Exploring Retention and Attrition in a Two Year Public Community College Retrieved from ERIC database. ( ED382257) Monk s, J. (2009). Who are the part time faculty? Academe, 95 (4), 33 37. Newman, J. E. (1977). Development of a measure of perceived work environment (PWE). The Academy of Management Journal, 20 (4), 520 534. Nonis, S. A., & Hudson,G. I. (2006). Academic perfo rmance of college students: Influence of time spent studying and working. Journal of Education for Business, 81 (3), 151 159. Organization for Economic Cooperation and Development [OECD]. ( 2004 ) Education at a Glance 2004 Retrieved from http://www.oecd.org/unitedstates/33714494.pdf Paccagnella, O.(2006). Centering or not centering in multilevel models? The role of the group mean and the assessment of group effects. Evaluation Review, 30 (1), 6 6 85. Pascarella, E. T. (1999). New Studies Track Community College Effects on Students. Community College Journal 69 ( 6 ) 8 14. Pascarella, E. T., Bohr, L., Nora, A., Desler, M., & Zusman, B. (1994). Impacts of on campus and off campus work on first year cognitive outcomes. Journal of College Student Development, 35, 364 370. Pascarella, E. T., & Terenzini, P. T. (1991). How college affects students San Francisco: Jossey Bass. Pascarella, E. T., & Terenzini, P. T. (2005). How college affects students: A third decade of research Volume 2. San Francisco: Jossey Bass. Paulsen, M. B. (1996). Higher education and state workforce productivity. Thought & Action, 12 (1), 55 77. Perna, L. W. (2005). The benefits of higher education: Sex, Racial/Ethnic, and socio economic group differences. Review of Higher Education, 29 (1), 23 52.
108 Perna, L. W. (2006). Understanding the relationship between information about college related behaviors. American Behavioral Scientist, 49 (12), 1620 1635. Peter, K & Horn, L. (2005). Gender differences in participation and completion of undergraduate education and how they have changed over time. Postsecondary education descriptive analysis reports. National Center for Education Statistics (ED), Washington, DC., MPR Associates, Berkeley, CA. Peugh, J. L., & Enders, C. K. (2004). Missing data in educational research: A review of reporting practices and suggestions for improvement. Review of Educational R esearch 74(4), 525 556. Ph illips, K. D & Campbell, D. F. (2005). Enculturation and development of part time faculty. In Wallin, D. L. Adjunct faculty in community colleges: an academic administrators guide to recruiting, supporting 53 71 Boston : Anker Publishing Company, Inc. Planty et al. (2007). The condition of education 2007 Washington, D.C.: U.S. Dept. of Education, National Center for Education Statistics. Retrieved from http://nces.ed.gov/pubs2007/2007064.pdf Planty et al.(200 8). Community Colleges: Special Supplement to The Condition of Education 2008 Washington, D.C.: U.S. Dept. of Education, National Center for Education Statistics. Retrieved from http://nces.ed.gov/pubs 2008/2008033.pdf Planty et al. (2009). The condition of education 2009 Washington, D.C.: U.S. Dept. of Education, National Center for Education Statistics. Retrieved from http://nces.ed.gov/pubs200 9/2009081.pdf in the U.S. News and World Report college ranking. International Journal of Educational Advancement 1 (2), 10 30. Price, J. L. (1977). The study of t urnover Ames: Iowa State University Press. Provasnik, S., & Planty, M ( 2008). Community colleges: Special supplement to the condition of education 2008 Washington, D.C: U.S. Dept. of Education, Institute of Education Sciences, National Center for Educa tion Statistics. Raudenbush, S.W. & Bryk, A.S. (2002). Hierarchical Linear Models (Second Edition). Thousand Oaks: Sage Publications. Reason, R. D. (2009). An examination of persistence research through the lens of a comprehensive conceptual framework. Jo urnal of College Student Development, 50 (6), 659 682.
109 Reason, R. D., Terenzini, P. T., & Domingo, R. J. (2007). Developing social and personal competence in the first year of college. The Review of Higher Education, 30 (3), 271 299. Reisberg, L. (1999 ). C olleges struggle to keep would be drop outs enrolled. The Chronicle of Higher Education 46(7), A54 56. Riegg, S. K. (2008). Causal inference and omitted variable bias in financial aid research: Assessing solutions. The Review of Higher Education, 31 (3), 3 29 354. Riegle Crumb, C. (2006). The path through math: Course sequences and academic performance at the intersection of race ethnicity and gender. American Journal of Education, 113 (1) 101 Riggert, S. C., Boyle, M., Petrosko, J. M., Ash, D., & Rude P arkins, C. (2006). Student employment and higher education: Empiricism and contradiction. Review of Educational Research, 76 (1), 63 92. Rose, A.S. & Hill, C. (2013). Women in Community Colleges: Access to Success Retrieved from https://aawcc.memberclicks.net/assets/documents/community%20colleges%20re port_2013.pdf Rosenbaum, J. E., Deil Amen, R., & Person, A. E. (2006). After admission: From college access to college success New York: R. Sage Foundation. Rosser, V. J., & Townsend, B. K. (2006). Determining public 2 year college faculty's intent to leave: An empirical model. The Journal of Higher Education, 77 (1), 124 147. Rubin, D. B. (1987). Multiple imp utation for nonresponse in surveys New York: Wiley. Ryan, J. F. (2004). The relationship between institutional expenditures and degree attainment at baccalaureate colleges. Research in Higher Education, 45 (2), 97 114. Sabharwal, M., & Corley E.A. (2009) Faculty job satisfaction across gender and discipline. Social Science Journal, 46 (3), 539 556. Sandy, J., Gonzalez, A., & Hilmer, M. J. (2006). Alternative paths to college completion: Effect of attending a 2 year school on the probability of completing a 4 year degree. Economics of Education Review, 25 (5), 463 471. Schuetz, P. (2005). UCLA community college review: Campus environment: A missing link in studies of community college attrition Community College Review 32(4), 60 80
110 Schuster, J. H. (2003 ). The Faculty Makeover: W hat Does It Mean for Students? New Directions for Higher Education 123 15 22. Scott, M., Bailey, T., & Kienzl, G. S. (2004). Institutions afloat? Determinants of college completion rates with a focus on the public urban college New York: Columbia University, Teachers College, Community College Research Center. Scott, M., Bailey, T., & Kienzl, G. (2006). Relative success? determinants of college graduation rates in public and private colleges in the US. Research in Higher Educa tion, 47 (3), 249 279. Scott Clayton, J. (2011). On money and motivation A quasi experimental analysis of financial incentives for college achievement. Journal of Human Resources, 46 (3), 614 646. Singell. (2002) .Merit, need, and student self selection: Is there discretion in the packaging of aid at a large public university? Economics of Education Review 21 (5) 445 454 Snyder, T., & Dillow, S ( 2012). Digest of Education Statistics 2011 Washington, D.C.: U.S. Dept. of Education, National Center for Edu cation Statistics. Retrieved from http://nces.ed.gov/pubs2012/2012001.pdf Snyder, T.D., Tan, A.G., and Hoffman, C.M. (2006). Digest of Education Statistics 2005 U.S. Department of Education, Nationa l Center for Education Statistics. Washington, DC: U.S. Government Printing Office. SSI. (2013). HLM Hierarchical Linear and Nonlinear Modeling (HLM). Retrieved from http://www.ssicentral.com/hlm/in dex.html Stampen, J. O. & Cabrera, E. A. (1986). Exploring the effects of student aid on attrition. Journal of Student Financial Aid 16 ( 28 37 ) Stater, M. (2009). The impact of financial aid on college GPA at three flagship public institutions. America n Educational Research Journal, 46 (3), 782 815. St. John, E. P. (1989). The influence of student aid on persistence. Journal of Student Financial Aid 19 (3), 52 68. St. John, E. P., & Paulsen, M. B. (2001). The finance of higher education: Implications f or theory, research, policy and practice. In M. B. Paulsen & J. C. Smart (Eds.), The finance of higher education: Theory, research, policy and practice (pp. 545 568). New York: Agathon Press.
111 Svanum, S., & Bigatti, S. M. (2006). The influences of course ef fort and outside activities on grades in a college course. Journal of College Student Development, 47 (5), 564 576. Swail, W. S. (1996). The development of a conceptual framework to increase student retention in science, engineering, and mathematics progra ms at minority institutions of higher education ( Doctoral dissertation ). Retrieved from ProQuest Dissertations and Theses (UMI No. 9703727) Swail, W.S. (2004). The art of student retention. Texas higher education coordinating board Paper presented at the 20 th Annual Recruitment and Retention Conference. Austin, TX. Swail, W.S., Redd, K. & Perna, L. (2003). Retaining minority students in higher education: A framework for success. In A. Kezar (Ed.), ASHE ERIC Higher Education Report 30 ( 2). Hoboken, NJ: Wil ey Periodicals, Inc. Terenzini, P. T., Pascare lla, E. T., & Blimling, G. S. (1996). Students' out of class experiences and their influence on learning and cognitive development: A literature review. Journal of College Student Development, 40(5), 610 23. Te renzini, P. T., & Reason, R. D. (2005). Parsing the first year of college: A conceptual framework for studying college impacts Paper presented at the Annual Meeting of the Association for the Study of Higher Education, Philadelphia, PA Terenzini, P. T., Yaeger, P. M., Pascarella, E. T., & Nora, A. (1996). Work study Paper presented at the meeting of the Association for Institut ional Research, Albuquerque, NM. Thirolf, K. Q. (2013). How faculty identity discourses of community college part time faculty change over time. Community College Journal of Research and Practice, 37 (3), 177 184. Thompson, E. R., & Phua, F. T. T. (2012). A brief index of affective job satisfaction. Gro up & Organization Management, 37 (3), 275 307. doi:10.1177/1059601111434201 Tin to, V. (1975). Dropout from higher education: A theoretical synthesis of recent research. Review of Educational Research, 45 (1), 89 125. Tinto, V. (2004). Student retention and graduation: Facing the truth, living with the consequences Retrieved from http://inpathways.net/ipcnlibrary/ViewBiblio.aspx?aid=2035 Tinto, V. (2006). Research and practice of s tudent retention: what next ? Journal of College Student Retention 8(1), 1 19.
112 Titus, M. A. (2004). An examination of the influence of institutional context on student persistence at 4 year colleges and universities: A multilevel approach. Research in High er Education, 45 (7), 673 699. Titus, M. A. (2006). Understanding college degree completion of students with low socioeconomic status: The influence of the institutional financial context. Research in Higher Education, 47 (4), 371 398. Townsend, B. K., & T wombly, S. B. (2001). Community colleges: Policy in the future context Westport, Conn: Ablex Pub. Corp. Townsend, B. K., & Twombly, S. B. (2007). Community college faculty: Overlooked and undervalued (ASHE Higher Education Report, Vol. 132, No. 6). San F rancisco, CA: Jossey Bass. Twombly, S., & Townsend, B. K. (2008). Community college faculty what we know and need to know. Community College Review, 36 (1), 5 24. Tyson, W. (2011). Negative impact of employment on engineering student time management, time to degree, and retention: Faculty, administrator, and staff perspectives. Journal of College Student Retention: Research, Theory and Practice, 13 (4), 479 498. Umbach, P. D. (2007). How effective are they? Exploring the impact of contingent faculty on unde rgraduate education. Review of Higher Education, 30 (2), 91 123. Umbach, P. D. (2008). The effects of part time faculty appointments on instructional techniques and commitment to teaching Paper presented at the 33rd Annual Conference of the Association fo r the Study of Higher Education. Jacksonville, FL Umbach, P. D., & Wawrzynski, M. R. (2005). Faculty do matter: The role of college faculty in student learning and engagement. Research in Higher Education, 46 (2), 153 184. U.S. Department of Education (2011). Beginning postsecondary students (BPS) longitudinal study (BPS:04/09). National Center for Education Statistics, Institute of Education Sciences. Washington, DC: Author. Van Landeghem, G., Onghena, P., Van Damme, J., & Opdenakker, M. (1999). The ef fect of different centering methods in multilevel analysis. Current Issues in Statistics, 53. Vartanian, T. P., Karen D., Buck, P., & Walker, C. (2007). Early factors leading to college graduation for Asians and non Asians in the United States Sociologi cal Quarterly the Sociological Quarterly, 48 (2), 165 197.
113 Wayman, J. C. (2003). Multiple imputation for missing data: What is it and how can I use it. Paper presented at the Annual Meeting of the American Educational Research Association, Chicago, IL We i, C. C., Berkner, L., He, S., Lew, S., Cominole, M., & Siegel, P. (2009). 2007 08 national postsecondary student aid study (NPSAS: 08): Student financial aid estimates for 2007 08. First look. NCES 2009 166. National Center for Education Statistics, Weis s, C., & Pankin, R. (2011). Part time faculty in higher education: A selected annotated bibliography. Retrieved from http://digitalcommons.providence .edu/cgi/viewcontent.cgi?article=1000&context= sociology_fac Wild, L., & Ebbers, L. (2002). Rethinking student retention in community colleges. Community College Journal of Research & Practice 26(6), 503 519. Williams Chehmani, A. (2009). The Effect of Fa culty Student Interactions on Student Engagement at Community C olleges (Doctoral dissertation ). Retrieved from ProQuest Dissertation and Theses database. (UMI No. 33515448) Wine, J. (2013). 2004/09 beginning postsecondary students longitudinal study: 04/09 methodology report Retrieved from http://nces.ed.gov/pubsearch/pubsinfo.asp?pubid=2012246 Winston, G.,& Zimmerm an, D.(2004).Peer effects in higher education. In C. Hoxby(Ed .), Coll ege choice: The economics of where to go, when to go, and how to pay for it Chicago: University of Chicago Press. Wong, G. Y., & Mason, W. M. (1985). The hierarchical logistic regression model for multilevel analysis. Journal of the American Statistical A ssociation, 80 (391), 513 524. Zwick, R., & Sklar, J. C. (2005). Predicting college grades and degree completion using high school grades and SAT scores: The role of student ethnicity and first language. American Educational Research Journal, 42 (3), 439 46 4.
114 BIOGRAPHICAL SKETCH Hongwei Yu was born and raised in a rur al community in Hebei, China. He attended Fengnan Secondary High School, Huazhong A gricultural U niversity (Bachelor of Arts and Master of Arts ). He received his Ph.D. in Higher Education Admi nistration from University of Florida in the August of 2013. He believe s higher education was, is, and will be the great equalizer for students, especially for students who come fro m disadvantaged backgrounds. He has great interest in studying community co lleges as these institutions of higher education are the great equalizers for students who want to make a difference in their lives through attending higher education.