INFLUENTIAL OR NOT: AN EXAMINATION OF THE INFLUENCE OF SOCIAL NETWORKS ON THE COLLEGE CHOICE PROCESS AND POST SECONDARY ENROLLMENT RATE OF LATINA/O STUDENTS By GRISELDA FLORES A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL O F THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2014
2014 Griselda Flores
To my nieces and nephew ( s ) : reme but based on everyone else who made it because of you
4 ACKNOWLEDGMENTS To my advisor and chair Dr. Bernard Oliver t hank you for your continuous support and dedication t hroughout this tedious process, especially during the final semesters. Next, I would like to thank my dissertation committee members: Drs. David Miller, Maria Coady, and Eileen Oliver. Thank you for your support and guidance from the early stages of my qualifying exam to the completion of this dissertation. Additionally, the knowledge I gained in some of your classes and outside of the classroom has allowed me to develop into an informed researcher and also provide d me with a strong foundation to be succe ssful in my future endeavors. To my formal professor, Dr. Luis Ponjuan, I would like to thank you for your support and dedication in my development as a scholar and researcher during the first two years of my journey as a PhD student. As a result of our numerous talks and writing sessions I am leaving this program confident in my abilities to conduct research and make a difference in the Latina/o community. Throughout the last several years I ha ve ha d the opportunity to call the University of Florida home. I am thankful for the many frien ds, colleagues, and classmates that have supported me throughout my doctoral program In particular, I am thankful for the ladies of 1215 Norman Hall for their ability to answer any question I posed to them especially Angela Rowe Additionally, my friends and classmates have been readily available to provide input, guidance, and support when I needed it most I am especially constant encouragement and endless guidance and for making this journey more tolerable and memorable these past four years Finally, I am so fortunate to have been a recipient of the Gates Millennium Y oung
5 Scholarship, and for there continues support not only financially but professionally and intellectually. To the Gates Millennium Scholarship, thank you for the various opportunities that have helped me become a more skill scholar and activist, and for the various mentors and friends I have meet as a result to these opportun ities and life changing experiences. I am truly blessed to come from a loving family that has constantly support ed and believed in me even when I doubted myself. To my parents, thank you for everything you have sacrifice for me, for your unconditional lov e and reassuring words of faith. To my four siblings, t hank y ou for all you have done for me. I know you were all just a phone call away To my big sister, Anita, thank you for being a loyal friend, a great listener, and for supporting my every move. To Ca rlos, Jay Jay, and Vale, I know if I ever needed anything I could always count on the three of you. As for my nieces and nephe w whose crazy personalities have made life much more enjoyable and memorable, my wish for you all is that someday yo u all will rea ch you r full human potential. Lastly, t o all my family members, close friends, and coworkers, thank you all for your inspiring words, you all made this journey worth taking.
6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 9 LIST OF FIGURES ................................ ................................ ................................ ........ 12 ABSTRACT ................................ ................................ ................................ ................... 13 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 15 Purpose of Study ................................ ................................ ................................ .... 21 Research Questions ................................ ................................ ............................... 22 Significance of Study ................................ ................................ .............................. 23 Definitions of Terms ................................ ................................ ................................ 25 Chapter Summary ................................ ................................ ................................ ... 28 2 LITERATURE REVIEW ................................ ................................ .......................... 29 Undergraduate Latina/o Students ................................ ................................ ........... 31 Demographic Characteristics ................................ ................................ ........... 31 College Enrollment Behavior of Latina/o Students ................................ ........... 35 First Generation College Students ................................ ................................ ... 38 Characteristics of Latina/o Students: Review o f the Literature ................................ 40 Latina/os Migration Patterns ................................ ................................ ............. 41 Social and Cultural Capital ................................ ................................ ............... 46 English Proficiency ................................ ................................ ........................... 49 Introduction to College Choice Process ................................ ................................ .. 50 Early Theoretical Frameworks ................................ ................................ .......... 51 Development of College Choice Models ................................ .......................... 52 ................................ ..................... 55 Layer One ................................ ................................ ................................ ......... 56 Layer Two ................................ ................................ ................................ ......... 56 Layer Three ................................ ................................ ................................ ...... 57 Layer Four ................................ ................................ ................................ ........ 58 College Choice Process of Latina/o Students ................................ ......................... 61 Predisposition ................................ ................................ ................................ ... 61 Search ................................ ................................ ................................ .............. 68 Choice ................................ ................................ ................................ .............. 72 Summary of Theoretical Frameworks ................................ ................................ ..... 77 3 M ETHODOLOGY ................................ ................................ ................................ ... 79 Proposed Hypotheses ................................ ................................ ............................. 80
7 Data Source ................................ ................................ ................................ ............ 80 Sampling Des ign ................................ ................................ ................................ ..... 82 Data Sample ................................ ................................ ................................ ........... 83 Descriptive Statistics ................................ ................................ ............................... 84 Dependent Variab les ................................ ................................ .............................. 90 Independent Variables ................................ ................................ ............................ 95 Layer One: The Individual Habitus ................................ ................................ .......... 97 Demographic Characteristics ................................ ................................ ........... 97 Social Capital ................................ ................................ ................................ ... 99 Cultural Capital ................................ ................................ ............................... 100 Supply of Resources ................................ ................................ ...................... 103 Expected Costs ................................ ................................ .............................. 103 Layer Two: School and Community Context ................................ ......................... 104 School Location Urbanization ................................ ................................ ......... 104 High School Type ................................ ................................ ........................... 106 Layer Three: Higher Education Context ................................ ................................ 107 Institutional Characteristics ................................ ................................ ............. 107 Methodological Considerations ................................ ................................ ............. 108 Analyzing Secondary Datasets ................................ ................................ ....... 108 Estimating Causal Effect with Secondary Data ................................ .............. 110 Analytical Methods ................................ ................................ ................................ 111 Limitation of Study ................................ ................................ ................................ 113 Contributions of Study ................................ ................................ ........................... 114 4 DATA ANALYSIS AND RESULTS ................................ ................................ ........ 116 Advanced Data Analysis ................................ ................................ ....................... 116 ........... 116 Multilevel Logistic Regression: Search Stage in 10 th grade ............................ 119 Multilevel Logistic Regression: Search Stage in 12 th grade ............................ 124 Multilevel Multinomial Logistic Regression: Choice Stage .............................. 128 Logistic Regression: College Enrollment ................................ ........................ 130 Chapter Summary ................................ ................................ ................................ 131 5 DISCUSSION OF FINDINGS ................................ ................................ ............... 145 Purpose of Study Revisited ................................ ................................ ................... 145 Summary of Research Findings ................................ ................................ ............ 147 Demographic Characteristics ................................ ................................ ......... 147 Cultural Capital ................................ ................................ ............................... 150 Social Capital ................................ ................................ ................................ 154 Supply of Resources ................................ ................................ ...................... 160 Expected Cost ................................ ................................ ................................ 163 School Context ................................ ................................ ............................... 1 66 Final Reflection of Findings ................................ ................................ ................... 167 6 IMPLICATIONS AND FUTURE R ESEARCH ................................ ........................ 181
8 Implications for Policy and Institutional Practices ................................ ................. 182 Institutional Practices and Recommendations ................................ ................ 183 Partnership between Spanish Language Organizations ................................ 189 Future Research ................................ ................................ ................................ ... 190 S ocial Networks ................................ ................................ .............................. 191 Gender, College Type, and College Applications ................................ ........... 192 Methodological Modifications ................................ ................................ ......... 193 Closing Words ................................ ................................ ................................ ...... 195 REFERENCES ................................ ................................ ................................ ............ 197 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 219
9 LIST OF TABLES Table page 3 1 Frequency of gender ................................ ................................ .......................... 84 3 2 Frequency of family income levels ................................ ................................ ...... 84 3 3 Frequency of family composition ................................ ................................ ........ 85 3 4 Frequency of number of in home siblings ................................ ........................... 85 3 5 ................................ ............................. 86 3 6 Frequency of parents English fluency ................................ ................................ 86 3 7 al status ................................ ........................ 87 3 8 Frequency of type of school ................................ ................................ ............... 87 3 9 Frequency of school urbanicity ................................ ................................ ........... 87 3 10 ................................ ............................. 88 3 11 ................................ .................. 89 3 12 Frequency of type of college to attend ................................ ................................ 89 3 13 Frequency of schools applied to ................................ ................................ ......... 90 3 14 Summary of dependent variables ................................ ................................ ....... 91 3 15 ............................ 92 3 16 Frequency of college search information (n= 1204) ................................ ........... 93 3 17 Frequency of college characteristics ................................ ................................ .. 94 3 18 Frequency of college enrollment (n=2209) ................................ ......................... 95 3 19 Summary of independent variables and indices ................................ ................. 95 3 20 Summary of multivariate logistic regression models ................................ ......... 115 4 1 aspirations ................................ ................................ ................................ ........ 133 4 2 Results of multilevel logistic regression model on 10 th stage ................................ ................................ ................................ ................. 134
10 4 3 Results of multilevel logistic regression model on 10 th stage ................................ ................................ ................................ ................. 135 4 4 Results of multileve l logistic regression model on 10 th stage ................................ ................................ ................................ ................. 136 4 5 Results of multilevel logistic regression model on 10 th stage ................................ ................................ ................................ ................. 137 4 6 search stage ................................ ................................ ................................ ..... 138 4 7 Results of multilevel logistic regression model on 12th gr search stage ................................ ................................ ................................ ..... 139 4 8 search stage ................................ ................................ ................................ ..... 140 4 9 choice stage ................................ ................................ ................................ ..... 141 4 10 choice stage ................................ ................................ ................................ ..... 142 4 11 choice college ................................ ................................ ................................ ... 143 4 12 Results of logistic ............... 144 5 1 aspiration and gender ................................ ................................ ....................... 145 5 2 ................................ ........................ 172 5 3 Results for multilevel ordi aspiration and number of schools applied to ................................ .................... 172 5 4 aspiration an d type of college ................................ ................................ ........... 172 5 5 aspiration and when students plan to enroll in college ................................ ..... 172 5 6 search choice ................................ ................................ ................................ ... 173 5 7 Results for multilevel logistic regression model on 10th g search stage ................................ ................................ ................................ ..... 174
11 5 8 search stage ................................ ................................ ................................ ..... 175 5 9 search stage ................................ ................................ ................................ ..... 176 5 1 0 search stage ................................ ................................ ................................ ..... 177 5 11 choice stage ................................ ................................ ................................ ..... 179 5 12 Results for binary log ... 179 5 13 Summary of significant independent variables for each multivariate logistic regression model ................................ ................................ .............................. 180
12 LIST OF FIGURES Figure page 2 1 ......................... 59 2 2 modified conceptual model ................................ ........................ 60
13 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 INFLUE NTIAL OR NOT: AN EXAMINATION OF THE INFLUENCE OF SOCIAL NETWORKS ON THE COLLEGE CHOICE PROCESS AND POST SECONDARY ENROLLMENT RATE OF LATINA/O STUDENTS By G riselda F lores May 2014 Chair: Bernard Oliver Major: Higher Education Administration This stud y investigat es the role of social networks on the college choice process and postsecondary enrollment of Latina/o students. The data source for this study came from the Educational Longitudinal Study of 2002 administered by the National Center for Educatio theoretica l frameworks for this study. Descriptive statistics and multivariate logistic regression techniques were utilize d to examine the influence of individual factors and e xpectations of social networks on college en rollment rate based on a nationally representative sample. The results suggest ed t hat particular variables in college enrollment. Namely, Latina/o students who applied to mul tiple college s ha d parents who were not fluent English speakers, plan ned to enroll in a vocational school, and anticipate d postposing college were less likely to aspired for a college degree.
14 Additionally, women were more likely to seek college informatio n from their guidance counselor W hereas Latina/o males and students who se native E nglish, who applied to on ly on e college, plan ned t o attend a vocational school, were first generation a nd who plan ned to postpone college were more likely n ot to seek college information from any source. Findings from this study highlight that the Latina/o student population is not a monolithic group but rather a heterogeneous population. Researchers and administrators must recognize this and develop poli cies that meet the diverse characteristics present ed among Latina/o student s Furthermore, high schools should implemen t college preparation programs and reach out to the Latina/o community in order to create a college going culture for this population Fi nally, the role of guidance counselors, teachers, and parents should be further examined with a heightened focus on understanding how their expectations for Latina/o students differ from other racial groups
15 CHAPTER 1 INTRODUCTION As the United States continues to change demographically, Latina/o students are becoming one of the dominant groups in K 12 classrooms, as well as in higher education. In 2007, Latina/os under the age of 18 became the second largest student group (Kohler & Lazarin, 2007 ). Given the fact that Latina/os now comprise the largest minority group in the United States surpassing African Americans, 16% vs. 13%, respectively (Ennis, Rios Vargas, & Albert, 2011), more national attention should be given to the underrepresentati on of Latina/o students in postsecondary education. Solorzano, Villalpando, & Oseguera (2005) captured the magnitude of the education in elementary school, only 52 will graduate from high school, of those students, only 10 will graduate from college, 4 of which will continue to graduate school, and only a Unfortunately, although Latina/os who will c omprise 24.4% of the U.S. population by 2030, it has been predicted that they will continue to have lower rates of degree attainment when compared to their White counterparts (Ennis, Rios Vargas, & Albert, 2011). In 2007, Latina/os comprised 18% of the tot al college age population. However, they only made up 10.4% of the graduating class of 2002 (Kohler & Lazarin, 2007). With college degrees replacing high school diplomas as the standard minimal attainment which is required for economic self sufficiency and responsible citizenship (Kuh, Cruce, Shoup, Kinzie, & Gonyea, 2008), it is essential that educators seek to understand the college choice process of Latina/o high school students. This would be an important
16 step towards assuaging the social and economic p roblems that will result from allowing Latina/o youth to grow into an undereducated and unskilled population. inequalities continue to exist for this population. According to the Ce nsus Bureau to 29.9% of White Americans. Although a growing number of Latina/o students are enrolling in college (25.8%), many more are dropping out without completing the ir Perason, 2006; NCES, 2009). Between 2000 and 2008, the enrollment of Latina/o students in degree granting institutions doubled from 1.4 million to 2.1 million, while White student enrollment increased by 14%, from 9 million to 10.3 million (Snyder & Dillow, 2009). W ith over half (54%) of college age Latina/os attending two year institutions, as compared to 25% of White students (Fry, 2002), there is still great room for impr ovement. Community college enrollment rates are much greater for Latina/o students over the age of 24. The issue with so many Latina/o students enrolling in community colleges is there retention and graduation rates, especially when it comes to transferrin g to four year institutions (Fry, 2002). Overall, Latina/os are more likely than school part time, enrolling in two year institutions, and/or delaying their college ed ucation into their mid Another group that has made great progress in college in recent decades are first generation students (Choy, 2001; Strayhorn, 2006). According to Saenz and associates (2007), one in six first year studen ts at four year institutions are considered
17 first generation. However, this number does not take into account first generation students enrolled at two accurate representation (30%) of first gener ation student enrollment rates across all postsecondary institutions, not only four year institutions. First generation students, like Latina/o students, are more likely to attend two year institutions (Chen, 2005), which is a critical factor for researche rs to consider when examining the college choice process and college enrollment decisions of first generation Latina/o students. It is also critical that researcher consider not only the type of institution (two year versus four year), but also the selecti vity of the institution, as both factors are influential in key outcomes such term social opportunities (Alon & Tienda, 2005). The economic value of a college education to Latina/os, parti cularly first generation Latina/o students, should not be underestimated. After all, the differences between college graduates and high school graduates continue to be propagated through annual earnings and employment opportunities. In 2009, the average me dian high school graduates (U.S. Bureau, 2012). Accordingly, Latina/o college graduates are likely to earn a higher salary when compared to those with only a high school d iploma. compared to a high school diploma (Julian & Kominski, 2011). Even when taking student loans into consideration, significant differences in lifetime earnings remain (Baum et al., 2010). In addition to earning a higher salary, a college education is also correlated with employment opportunities (Institute for Higher Education Policy, 2005).
18 According to the U.S. Bureau of Labor Statistics (2012), workers over th e age of 25 with only have a high school diploma who are twice as likely (9.2%) to be unemployed. Due o new settings, substantial benefits that result from an investment in higher education. Moreover, an educated Latina/o workforce with higher earnings and a lower unemployme nt rate will benefit the United States as a whole. Within the past decades, the growth within the Hispanic population has transformed and will continue to influence Latina/ o population increased from 35.3 million in 2000 (when this group only comprised 13% of the total U.S. population) to 50.5 million in 2010 or 16% of the total U.S. population a 3% increase (Ennis, Rios Vargas, & Albert, 2011). Although Latina/os continue to concentrate in traditionally Latina/o states such as, Texas and California, there has been a rapid increase in nontraditional states in the Midwest and Southeast (Kohler & Lazarin, 2007). As of 2010, 41% of the Latina/o population lived in the West, 36 % lived in the South, 14% in the Northeast, and 9% in the Midwest (Ennis, Rios Vargas, & Albert, 2011). The two states with the largest Latina/o population, Texas and California, have invested a great deal of money and resources in programs and research th rates in their respective states. While these two states have taken some initiatives to deal with the Latina/os educational crisis, many more states are not prepared to deal with the rapid influx and
19 growth of Latina/os within their states. For example, the Hispanic population more than doubled in eight states in the South: Alabama, Arkansas, Kentucky, Maryland, Mississippi, North Carolina, South Carolina, and Tennessee (Ennis, Rio s Vargas, & from 5% to 17%, in the Mideast from 2% to 7%, in the Midwest from 2% to 7%, and in f South Carolina experienced the fastest growth, increasing from 95,000 in 2000 to 236,000 by 2010 (Ennis, Rios Vargas, & Albert, 2011). There is an urgent need for every state to ensure all their residents understand the long term benefits of higher educa tion. Additionally, it is critical that state implement the necessary programs and policies in order to increase the educational attainment of Latina/o students within their respective state, and the country (Santiago, 2011). These are common sense measure s, as a college educated labor force that includes the participation of Latina/os is considered necessary for the growth of the U.S. domestic economy (Badger, 2010) and to assist the United States in competing in the global market (Hispanic Alliance, 2010) The educational attainment of Latina/os is also higher education degrees by 2020 (Gonzalez, 2010; Santiago, 2011). With the rise of technology in the past few decades, m ore jobs are transitioning away from manufacturing and towards a high tech economy (Badger, 2010) that requires some college education. Unfortunately, less than 20% of Latina/o workers are employed in high tech occupations (U.S. Bureau of Labor Statistics, 2009) with the vast majority concentrating in low skill occupations. The discrepancy that exists between the
20 education and skills of Latina/os and the education and skills required for the U.S. workforce to effectively compete in the global market, sugges ts increased attention to An important aspect of this greater issue is understanding that, for many Latina/o nzini et al., 1994, p.62) after high school. This tendency to not perceive college as the natural next step, in part accounts for the trend of so many Latina/o students entering college at a later age. Moreover, since many Latina/o students are the first i n their family to attend college, the decision to enroll in college often represents a significant departure from their background and experiences. Enrolling in college exposes one to new academic and social climates (Terenzini, et al., 1994) that are ofte n foreign to many first generation students and families. However, not all students experience the college choice process the same way (Glick & White, 2004, Kim, 2004). Factors such as race, ethnicity, and generational status have been reported to mediate, to a certain extent, college choice decisions and outcomes for students (Ceja, 2001; Hamrick & Stage, 2004; P.A. Perez, 2007). To better address the low college enrollment rates of Latina/o students, it is crucial to examine how these students make decisi ons about college. Two researchers identifying how Latina/os come to formulate postsecondary plans and navigate their college choice decisions, we can enhance their educatio achieving Latina/o high school students in California, further research is needed to determine whether the college choice
21 process is the same for all Latina/o students. In an effort to address this gap in the literature, the present study will focus on the college choice process of Latina/o students (both, first generation or non) who were sophomore students in 2001 2002 in multiple states across the United States. With the ultimate aim of identifying and understanding examined the process that Latina/o students experienced when making decisions about college. These factors were examined in relation to how the expectations of certain social networks influence their college choice process. Purpose of Study college enrollment rates, it is essential to understand those factors t hat influence their college choice process in order to strengthen the educational pipeline for Latina/os. With many Latina/o students delaying college enrollment or not enrolling at all, it is important to know who is impacting those few Latina/o students to aspire and enroll in college after high school. Of those few Latina/os who enroll in college, the majority enlist in two year institutions (Fry, 2002). As previously mentioned, the problem with enrolling in two year institutions is the low rates of rete ntion and college completion, especially among Latina/o students who often enroll part time, delay college enrollment, and/or have multiple family obligations to attend to (Fry, 2002). Although a substantial amount of research exists on those factors that influence the college choice process of both low income and students of color, very few studies focus exclusively on first generation Latina/o high school students. The purpose of this study is to examine how the college choice process of this population i s impacted by the
22 relatives, friends, etc.), as interpreted by the student, in addition to other variables. osition, search, choice stage, and college enrollment (Hossler & Gallagher, 1987). Descriptive statistics and multilevel logistic regression analyses were performed on data gathered from the Educational Longitudinal Study of 2002 (ELS:02) conducted by the U.S. Department of Education, Institute of Education Statistics. The ELS:02 dataset initially surveyed tenth grade students in 2002 and followed up biennially in 2004 and 2006. As the focus of the was used from all three Research Question s The research questions guiding this empirical study focus on identifying those social networks or individuals (i.e., guidance couns elors, teachers, parents, close relatives, or friends) who contributed the most to the college choice process of Latina/o students. This study is designed to explore four research questions that pertain to the college choice process of Latina/o students: Does the expectations social networks have for Latina/o high school students, along Does the expectations social networks have for Latina/o high school students, along with other key variab entrance information in both the 10 th and 12 th grade? Does the expectations social networks have for Latina/o high school students, along with other key variables, influence the primary reason stu dents selected their chosen college? Does the expectations social networks have for Latina/o high school students, along with other key variables, influence their decision to enroll in college? In examining the college choice process of Latina/o students there are various conceptual frameworks to consider. However, for the purpose of this study it was
23 important to find a framework that included relevant factors identified in the literature as influential in the college choice process of first generation students and Latina/o students. The decision to focus on first generation Latina/o students was primarily driven by the fact that a large percentage of Latina/o students are considered first generation college students. Additionally, this group is believed to face unique challenges above and beyond those experienced by students who have college educated parents. Two conceptual frameworks were selected based on their ability to 87) three of student college choice. The selected frameworks will be discussed in greater detail in the following chapters. It is important to note that this study focuses o n within group differences of a Latina/o student sample, in terms of income, parental education, study does not include a comparison between Latina/os and non Latina/o s on the variables of interest (Phinney, Dennis, & Chuateco, 2005). Significance of Study Moreover, this study is significant because it capitalizes on those factors that further contributes to and expands on the limited literature that focuses on those factors frequent migration patterns and high birth rates of Latina/os (1 of 5 school age children enroll in college is a social and economic imperative (San Francisco Chronicle, 2009). Unfortunately, Latina/os also account for 24.3% of the populatio n living in poverty
24 (Ramirez & de la Cruz, 2003). According to the U.S. Department of Health & Human Services (2013), the poverty line for a household of four is $23,550, with the average size of Hispanic households being larger than non Hispanic household s, 3.47 people compared to 2.62 for the total population (American Community Survey, 2008). In will become more problematic because by the year 2018 the United States will demand that 63% of the American workforce be college educated (NALO, 2010). As the Latina/o population continues to increase they will form a crucial segment of American workers, taxpayers, parents, citizens, voters, and leaders (Population Reference Bureau, 2010). It is projected that by 2050 one third of the US population will be Latina/o gap is not addressed the United States will have to pay a higher price. For e xample, the cost to US taxpayers for college dropouts among all first time freshmen between 2003 and 2008 was 9 billion dollars (American Institutes for Research, 2010). The United States will continue to suffer an economic decline if Latina/os do not reac h their potential human capital for an economy that will eventually require a baccalaureate degree from the majority of its workforce. I f the U.S. is to continue to compete in the global economy then education which has long been the gateway to enhanced pe rsonal and social economic security must be made more accessible to Latina/o students. This study is significant for many reasons and to a variety of stakeholders. For secondary and postsecondary administrators and educators, this study helps to better e lucidate which non academic factors and social networks are most influential and/or
25 enroll in college. For policymakers, this study will supply critical information regardi ng an generation Latina/o students. Armed with this information, policymakers can provided secondary and postsecondary institutions with the necessary funding to establish or support existing rates. In addition, this study will help to provide practical recommendations to gap through institutional policies, programs, and support. All in all, this study will contribute college choice process and postsecondary enrollment. It is critic al that policymakers and educational leaders acknowledge that Latina/o degree attainment is not only an individual issue but a national concern that contributes to a decline in the average income levels, lower average tax contributions, and a n increase in the number of unfulfilled professional jobs (Museus, Fry, 2002, p.1). At a similar r (Fry, 2002, p.1). It is imperative that educators and policy makers become better informed an d proactive in making the necessary decisions to help increase the college enrollment and graduation rates of Latina/o students. Definitions of Terms T here are various terms used throughout the literature that explain the college choice process of Latina/ o students. While many of these terms may be used
26 interchangeably, many others should not be. Therefore, the following terms are defined so the reader can better understand the research study at hand and avoid any misinterpretation of pertinent terminology Latina/os. government imposed label used since t he 1970s to identify anyone who spoke Spanish or was a descendant of Spaniards. However, fo r the purpose of this paper these two terms (Hispanic and Latina/o) will be used interchangeably to refer to both male and female natives or descents of Spanish speaking countries in South and Central America, including some of the Caribbean islands (i.e., Puerto Rico, Dominica n Republic, and Cuba). Latina/os are often referred to as a homogeneous group that share much in common however; significant within group differences exist among Latina/os in a variety of domains to include social and economic status, educational attainment, and political views. Nonetheless, one thing Latina/o descents share in common is their history of Spanish colonization and native language, Spanish. Due to the shared history of colonization among Latin American countries, Spaniard s are excluded as a Latino or Hispanic subgroup in th is study. First generation college student. student whose parents never enrolled in postsecondary education (Nunez & Cuccaro Alamin, 1998; Pascarella, Pierson Wolniak, & Terenzini, 2004). While other studies define first generation as undergraduates whose parents have never obtained a Mendez, 2003). Still others define first generat ion as a student whose parents never attended college nor have any older siblings who went to college before them (Rooney,
27 2008; Talavera e parents never obtained any form of postsecondary college degree (i.e., Associate of Arts/Science or Bachelor degree). College choice process. process high school students go through when deciding whether an d where to go to phrase model of college choice to get a better understanding of how students make decisions about college attendance. The model divides the college choice process into three sta ges: (1) predisposition, the decision to go to college, (2) search, searching for general information about college and learning about certain institutions; and (3) choice, completing applications and choosing an institution for enrollment (A.F. Cabrera & La Nasa, 2000a; Hossler et al., 1999). Traditional college student. For this study, this term will refer to full time, first seeking students, enrolled in college within six months following their high school graduation (Knapp et a l., 2007). Socioeconomic status (SES). A combination of factors, including family income, parental education, and occupational status of parents, will be used to define socioeconomic status (Harwell & LeBeau, 2010). Socioeconomic status is often highly cor students from lower socioeconomic status lack (Walpole, 2003). Cultural capital Characteristics including language, cultural knowledge, and manners that result from an indi class status (Bourdieu, 1986).
28 Social capital social network and the amount of social, economic, and cultural capital these networks poss ess (Bourdieu, 1986). Social Networks. Individuals who can directly transit or negotiate the college choice process (Stanto Salazar, 2004). However, for this study I will be referring to institutional gatekeepers, (i.e., school counselors, teachers, and athletic coaches), as process through personal advice, support, or institutional res ources. Chapter Summary Chapter 1 highlights the consequences of the increasing disparities in enrollment in postsecondary education of Latina/o students and their rapid demographic growth. As more Latina/o students aspire to enroll in college, it is crit ical to understand their college college choice process, among other variables. The purpose of this study was to examine the degree of influence certain social networks and non academic factors have as a heterogeneous group, attention was given to background differences amongst Latina/o students, such as parental education, native l anguage, and high school background. The Educational Longitude S tudy of 200 2 was utilized for this study because it provided the most comprehensive and recent source of national dataset
29 CHAPTER 2 LITERATURE REVIEW More researchers are studying the college choice processes of Latina/o students especially in regards to their decision to enroll in college or not (Ceja, 2006; Hurtado Ortiz & Gauvain, 2007; Rooney, 2008; P.A. Perez & McDonough, 2008). However, very few researchers have examined how the expectations of those individuals who may have the closest relationship with Latina/o students impact their college choice process. Learning more about this topic may help us to understand some of the dec isions Latina/o students make about the college choice process, especially since more than half (55%) of Latina/o students tend to enroll in two year colleges versus four year institutions (Adelman, Daniel, & Berkovits, 2003). As previously mentioned, Lati na/o students who start off at community colleges are less likely to transfer to a four year Laden, 2004). Attending college, more specifically four year institutions, has proven beneficial over the lifetime of the individual. $57,026, compared to $40,556 for those with so me college education, but no degree (Julian & Kominski, 2011). The median income in 2010 of $29,900 is much lower for individuals with only a high school diploma or its equivalent (U.S. Department of is becoming an essential job requirement for many up and coming occupations. Within the past decades, eight of the employment (U.S. Bureau of Labor Statistics, 2009). The need fo r more college
30 critical for the United States especially when it comes to competing in the global market. According to Adelman (2009), there is a growing concern that other co untries will surpass the United States in the proportion of bachelor degree recipients they educate. In addition, there are many other benefits that come from a college educated society: healthier citizens, reduction in crime rates, decline in unemployment rates, an improved infrastructure, less dependency on government assistance, and an increase in political participation and volunteerism (Baum & Ma, 2007). Thus, a more educated society translates to a more powerful and wealthier country for the United St ates as a whole. The purpose of this study is to examine the role that certain social networks or literature review, the first section will explore the demographic ch aracteristics of the Latina/o population within the United States and provide context for the theoretical framework of this study. The second section provides an overview of the research literature done on first generation and Latina/o college students. Th e overview includes studies that focus on the habitus (i.e., demographic characteristics, cultural capital, and ion, the development of various college choice models that have impacted current research studies will be reviewed. The final two sections will provide a review of the main theoretical frameworks used for this study, as well as an overview of the literatur e that focuses on the various process. This line of research will help us to further understand the college choice
31 process of Latina/o students, and their subsequent enroll ment or lack thereof, at institutions of higher education. Afterwards, a brief summary of the theoretical frameworks used for this study will be provided. These studies include Hossler & ceptual model. Undergraduate Latina/o Students The characteristics of Latina/o students focus on a few of the elements that may help or hinder their college choice process. As mentioned earlier, it is important to note that there are significant differen ces among Hispanic college students, which will be explored in further detail through the examination of their college choice process. However, the purpose of this overview is not to focus on the differences between U.S. born versus foreign born Latina/o s tudents but will provide a broad summary of all Latina/o undergraduate students in order to better comprehend their college choice process as a larger group. As with the differences in every other racial group, demographic characteristics will evolve (i.e. attainment, generational status, gender). Demographic Characteristics Even though they represent a substantially diverse ethnic group, the average Latina/o college student is American born, female, of non tradit ional college age, and of lower socioeconomic status. With recent media coverage of the Development, Relief, and Education for Alien Minors, or as its better known, the DREAM Act, (Mack, 2011; Navarrette, 2010; Perez, 2010) the dialogue on Latina/o student s in higher education has focused more attention on undocumented students. The proposed federal legislation (Dream Act) intends to provide undocumented students with the opportunity to attend college. Although the media has heavily covered undocumented stu dents
32 small percentage of the Latina/o undergraduate population. Of those Latina/o students in college, 88% are U.S. born citizens and another smaller percentage (11%) are legal representation in postsecondary institutions may have to do with their legal status and their inability to attend or pay for international college fees. While it is impor tant to address undocumented students in the literature, the database used for this study does not provide such information. As further explained in the next chapter, participants were asked about their generational status, as well as that of their parents (i.e., U.S. citizenship or immigrates), but not for their legal status (i.e., documented vs. undocumented). Therefore, this particular characteristic (i.e., legal residents/citizens or undocumented) cannot be controlled for in this study and instead was e xcluded. In addition to being U.S. born, the typical Latina/o college student is a female. Within the past years, the college enrollment of Latina/o students has steadily (S antiago, 2008). As of 2009, Latinas represented 57.8% of all Latina/o students in higher education (Saenz & Ponjuan). Some researchers such as Ceja (2001) and Talavera Bustillos (1998) have taken into consideration the influence of gender on the college c hoice process. For example, Ceja (2001) examined the college choice process and destinations of first generation Chicanas (i.e., U.S. born individuals of Mexican re convinc dimensions explaining wh y their
33 Talavera Bustillos found similar gender issues among first generation Chicanas, many of whom confronted expectations t o stay home and begin raising a family. While some researchers, like Ceja and Talavera Bustillos (Contreras Godfrey, 2009; Zarate & Gallimore, 2005; Saenz & Ponjuan, 2009) suggest that men and women experience the college choice process dif ferently, many others report that gender has little or no effect on the college decision making process of Latina/os (Hearn, 1991; Paulsen, 1990). Most recently, researchers like Saenz & Ponjuan (2009) have started to tion in higher education. Saenz and Ponjuan (2009) found that Latino males are less likely to enroll in college compared to Latinas due to the following factors: strong recruitment into the military, overrepresentation in the U.S. prison system, being labe and the expectation that they must provide financial support for their immediate family due to a sense of familismo and machismo presence. Familismo consists of maintaining fulfilling familial obligations, and holding strong beliefs about familial support (Sabogal, Marin, Otero Sabogal, Marin, & Perez Stable, 1987; Suarez Orozco & Suarez Orozco, 1995). Machismo, on the other hand, comprises a strong or exaggerated sense of m anliness, stressing attributes such as physical courage, virility, domination of women, and aggressiveness (Suarez Orozco & Suarez Orozco, 1995). Suarez Orozco & Suarez Orozco (1995) reported that feelings of duty and responsibility to the family could tak e priority over education, especially if the family is struggling financially. However, others have found familismo to be one of the most positive predictors of academic achievement (Portes, 1999).
34 Despite the increase in college enrollment of Latina/o st udents, the average Latina/o student is older than his/her peers. The traditional age of college students is between the ages of 18 to 24 (Synder & Dillow, 2009). Approximately 7% of Latina/o college students are of non traditional age compared to 5% of Wh ite students (Fry & Eric Clearinghouse on Urban Education, N.Y. 2003). When looking at different racial groups, only 37% of Latina/o high school graduates enroll in college versus 50% of White students and 40% of African American students (Snyder & Dillow, 2009). The issue with students delaying college enrollment is that this decreases their chances of According to the research findings of Horn et al. (2006), undergraduate students who delay college enrollment are more likely to enroll at a community college (56%) versus 34% of undergraduates who enrolled at a two year college after high school. After all, enrollment in a community college decreases the chances for any student, especially Lati 2006; Laden, 2004). It has been suggested that outcomes of students who delay their college enrollment should not be compared with those who enroll in college after high school since on e can argue that both groups differ in many aspects. Therefore, many college choice studies, including this study, have been careful to include only traditional college age students who enroll in college after high school and not non traditional college st udents (Bers, 2005; Rooney, 2008). Further details on this topic will be provided in Chapter 3 Moreover, the average Latina/o undergraduate student comes from a lower socioeconomic status. In particular, dependent Latina/o students tend to come from
35 fami lies with lower income levels in comparison to all undergraduates (Santiago, 2007). Dependent students consist of individuals who are required to provide parental information or signatures on the Free Application for Federal Student Aid (FAFSA) [Federal St udent Aid, 2010]. Santiago & Cunningham (2005) found that nearly 25% of dependent Latina/o students came from families with household incomes below $40,000, compared to about 21% of all undergraduates in 2003 2004. To better understand the college choice p rocess of low income Latina/o students, many researchers have narrowed their focus towards this particular group (Collatos, Morrell, Alejandro, & Lara, 2004; Oliverez, 2006; Rooney, 2008). Nonetheless, some researchers have argued that factors other than p arental income have been more influential in the college decision making process of Latina/o students (A.F. Cabrera & La Nasa, 2000b; McDonough, 1997). College Enrollment Behavior of Latina/o Students Giving the distinctive characteristics of Hispanic coll ege students is not enough without also highlighting their college enrollment behaviors (Fry, 2012; Santiago, 2007). The college attendance behaviors of Latina/o students examines whether the students are enrolled in two year or four year institutions and whether they are studying part time versus full time. According to data from the U.S. Department of Education (ED, IES, & NCES, 2010) 47% of Latina/o students were enrolled in college as part time students in 2009. Among the undergraduate student populatio n, Latina/o students are the least likely to be enrolled in college full time. For example, the part time enrollment rate of other racial groups in the study were 37% for Whites, 40% for Blacks, 37% for Asian/Pacific Islanders, and 39% for American Indian/ Alaska Natives (U.S. Department of Education ED, IES, & NCES, 2010).
36 Researchers such as Fry (2002) and Nora & Rendon (1990) have suggested that Hispanic students typically enroll in college part time to help support their families financially. Hearn (199 2) found that students from lower SES backgrounds are more likely to enroll in college part time than students from middle or upper class backgrounds. Researchers have argued that Latina/os who enroll part time differ characteristically from those who enro ll full time (Hearn, 1992; Terenzini & Pascarella, 1998). Unfortunately, the limited literature on the college choice process of Latina/o students focuses mainly on the ability of Latina/o student to successfully navigate the college choice process (Contre ras Godfrey, 2009; Talavera Bustillos, 1998). Hardly any distinctions were made about their college enrollment status. In other words, many postsecondary institution wit hout much consideration about whether they enrolled part time or full time, at either a four year or two year institution, all of which would be influential on the ability of the students to graduate from college. Nearly half of Latina/o students between the ages of 18 to 24 are enrolled in two year institutions (Fry, 2012). This is larger than any other racial group. For example, only 27% of Whites, 37% of Blacks, and 22% of Asian American are enrolled in two year institutions (Fry, 2012). Some researcher s have asserted that SES status (Kao year institutions versus four year institutions. Also, t he cost of tuition at community colleges is much lower than that of four year institutions, which explains why many low income students are attracted to two year institutions. However, SES alone is not
3 7 sufficient for explaining the high rates of enrollment of Latina/o students in community colleges. Admon (2006) found that when taking low SES into account for different racial groups, Latina/o students were still more likely than White or Black students to enroll in two year institutions. Admon (2006) sugges ted that the type of social capital possessed by Latina/o students often results in a lack of information about college costs and financial aid which may result in their higher enrollment rates in two year institutions compared to their White and African A merican counterparts. Limited literature on the college choice process of Latina/o students who are enrolled in four year institutions exist (Cohen, 2009; Contreras Godfrey, 2009; P.A. ollege choice process during high school (Anderson, 2008; Ceja, 2001; Carreras, 1998; Gomez, 2005; L. Gonzalez, 2007; Kao & Tienda, 1998; Oliverez, 2006; P. A. Perez, 2007; Wolf, 2007). As previously mentioned, many studies do not take into consideration t he type of institutions (i.e., two year or four year) students are more likely to enroll in after high school (Anderson, 2008; Carreras, 1998; Gomez, 2005; Kao & Tienda, 1998; P.A.Perez, 2007; Wolf, 2007). Although it seems reasonable for researchers to fo cus on Nasa, 2000a) some researchers have found an inconsistency between Latina/o sch complete the college choice process and actually enroll in college. This type of research predisposition, search, and choice stage, as well as their college enrollment decision.
38 As suggested by Swail, Cabrera, Lee and Williams (2005c), it is essential to follow in to what this study intends to do. Furthermore, Perna (2000) argues for more studies to take into account the type of institutions (i.e., two year or four year) stu dents intend to enroll in. While many studies have focused on the college choice process of Latina/o students who choose the community college route, more studies should focus on those Latina/os who have broken away from the norm and enrolled in four year institutions (Admon, 2006; Benitez choice process in four year institutions will better inform secondary and postsecondary administrators and policymakers on ways to impro ve the college enrollment rates of Latina/o students at four year colleges. While this study does not provided information on the type of institution Latina/o students actually enrolled in (i.e., two year college, four year college, or vocational) after hi gh school, it does provided information on the type of college Latina/o students aspire to attend after high school. This information may First Generation College St udents generation students mainly because most Latina/o students are the first in their family to attend college (Nunez & Cuccaro Alamin, 1998). An interest in first generation college college access and completion rates compared to students who have college educated parents (Ceja, 2001). This section presents a brief description of the demogra phic
39 characteristics of first generation college students. In particular, first generation college students at four year institutions which is the ideal population of interest for this study. Unfortunately, since the data used for this study does not menti on what type of institution, two year or four year, Latina/o students enrolled in after high school, this information is left unknown. In a series of studies conducted by the National Center for Education Statistics (NCES) on the experiences of high schoo l graduates and first generation college students (Choy, 2001), findings indicated that students who had parents who were college graduates had an advantage over first generation college students. Among that population of high school graduates, 59% of firs t generation students enrolled in college by 1994 in comparison to 93% of students whose parents had a college degree (Choy, 2001). Although lower parental education reduces the likelihood of students enrolling in college, NCES stresses that this is not th e only factor linked to college enrollment. Therefore, investigating first lessen the influence of parental education (Choy, 2001). Swail et al. (2005c) found support for the NCES hypothesis after disc overing that first generation Latina/o students educated parents once academic preparation and aspiration were controlled for. While the national average enrollment rate of first gene ration freshman students in 2005 was 16%, the proportion was much higher for Latina/o students at 38% and 23% for African Americans (Saenz et al., 2007). Whereas all racial and ethnic groups have shown a decline in the representation of first generation st udents, the proportion is the highest for Latina/o students and the lowest for White students (Saenz et al., 2007).
40 overrepresentation in community colleges and their lack of ac cess to four year colleges. Findings in the study by Saenz et al. (2007) reveals that first generation students are more likely to do the following: attend college because their parents want them to; work in order to pay for college; attend college to incr ease their income; and take into account financial factors when choosing a college. Unsurprisingly, first generation students are less likely to live on campus. They are also less likely to be academically self confident, especially in their self rating o f math and writing ability (Saenz et al., 2007) which impacts the type of institutions they choose to enroll in. The problem with first generation students not living on campus is th eir increase chances of not being retained. After all, research has shown that student involvement and living on success rate (Astin, 1973; Astin, 1999; Bozick, 2007; Tinto, 1994). This is particularly true of residence halls since many provide d students with academic sup port, leadership opportunities, and peer interaction, in addition to close proximity to campus, which helps to promote student retention (Li, Sheely, & Whalen, 2005). Unfortunately, with over half of Latina/o students enrolling at community colleges, the l ikelihood that they will live on campus is small since on campus housing is rare a mong two year institutions. Characteristics of Latina/o Students: Review of the Literature es sential to first discuss who Latina/os are, as well as their place in American society. A brief description of Hispanic immigration, social and cultural capital, and language context will be provided to understand how these contexts may or may not shape th e college choice process of Latina/o students. Without adequate knowledge about these
41 characteristics, practitioners and policymakers can make incompetent decisions concerning outreach, recruitment, admissions, and retention programs that affect this popul ation. In the literature, Latina/os are often grouped together and treated as a homogenous group who share similar histories, struggles, and experiences. Unfortunately, Latina/os vary in every aspect of the spectrum, with migration alone varying among each different Latina/o subgroup. As a result of the various migration United States, one can argue that the educational attainment of specific Latina/o subgroups have b een impacted, to a certain extent, by these foreign relationships. Latina/os Migration Patterns Mexican Americans account for over half (63%) of the Latina/o population in the United States (Ennis, Rios Vargas, & Albert, 2011). For generations, several Me xican families have lived in modern day U.S. states such as, Arizona, Utah, Nevada, California, Texas, New Mexico, and parts of Colorado and Wyoming which were ceded by the United States from Mexico in the 19 th century. However, the majority (58%) of Mexic ans arrived in the U.S. in the 1990s or later, with 37% arriving within a decade of the 1990s (U.S. Census Bureau, 2007). Even today, the majority of Mexican Americans reside in states located on the West coast, in particular California and Texas which acc ount for 61% of the Mexican American population (Ennis, Rios Vargas, & Albert, 2011). According to Portes and Rumbaut (2006), Mexican migration began in the early 20 th century due largely to the Mexican Revolution, and the recruitment of Mexican workers by U.S. growers and railroad companies. Today, Mexicans account for about 30% of the U.S. immigrant population, including naturalized citizens and noncitizens, regardless of their immigration status (U.S. Census Bureau, 2009).
42 During the 20 th century, the M exican government twice entered into an agreement with the United States to export Mexican workers as contract laborers while American soldiers fought overseas, in both World War I and WWII. However, increases in a nativist climate and an economic downturn in the U.S. resulted in the elimination of these contracts and the creation of the U.S. Border Patrol, along with other immigration laws which restricted the flow of Mexican immigrants into the U.S. (Durand & Massey, 1992). Eventually, this resulted in an increase in illegal immigration from Mexican migrants who entered into the United States. Today, 60% of all unauthorized immigrants to the United States come from Mexico and about 62% of all Mexican immigrants were unauthorized in 2009 (Passel & Cohn, 201 0). The majority of Mexicans migrated to the United States for economic reasons. They end up in a workforce which overwhelmingly consists of manual workers who possess limited skills and have low levels of schooling (Portes & Rumbaut, 2006). Consequently, educational attainment (Portes & Rumbaut, 2006). As previously mentioned, Latina/os do not all share the same migration history. For examp le, the second and third largest Latina/o subpopulations consist of Puerto Ricans (9%) and Cubans (4%) [Ennis, Rios Vargas, & Albert, 2011]. Unlike the rest of Latin America, Puerto Rico has been an unincorporated territory of the United States since 1898, and in 1917 Puerto Ricans were granted U.S. citizenship. Thus, Puerto Ricans have the luxury to travel back and forth without any migration issues. Puerto Ricans have been migrating to the states since the 19 th century. Nevertheless, Puerto Ricans have kept a strong hold on their Puerto Rican identity by preserving their cultural
43 heritage and their native language, Spanish, while simultaneously resisting American cultural influences (Brick, Challinor & Rosenblum, 2011). Contrary to Mexican Americans, who mainly reside in West ern states, the majority of Puerto Ricans have settled in New York, with a growing Puerto Rican community in central Florida (Ennis, Rios Vargas, & Albert, 2011). Idiosyncratically, Cubans also differ in regards to migration patterns. Although Cubans have been migrating to Florida since the Spanish colonial period, with smaller waves coming in the early 20 th century, a large influx of Cuban migration started during the post Castro revolution (Luis Brown, 1967; Mendez, 1994). After the Cuban revolution, led by Fidel Castro in 1959, which resulted in his association with the Soviet Union and the introduction to communism, thousands of Cubans left Cuba and migrated to the United States. The first wave of Cubans to migrate to the United States, after the post C astro revolution, consisted mostly of highly educated upper and middle class Cubans (Mendez, 1994). Contrary to other Hispanic immigrant groups, Cubans were well received by the United States government and admitted into the U S as political refugees. The U S anti communist sentiments and their opposition to the Soviet Union influenced the American government to establish many programs to assist Cuban refugees. For example, the Cuban Refugee Program established by the President of the United States in 196 1 provided direct financial assistance to Cuban refugees via public assistance, Medicare, free English courses, employment opportunities, scholarships, and low interest college loans (Mendez, 1994; Luis Brown, 1967). Many Cuban refugee doctors were also a llowed to take a refresher course offered at the University of Miami Medical School (Thomas, 1963) in order to regain
44 their medical credentials. Upon completion of this course and upon passing the Educational Council for Foreign Medical Graduates examinati on (ECFMG), these Cuban doctors continued to practice medicine under the supervision of an American doctor (Thomas, 1963). Many other Cuban professionals were given refresher courses, training, and educational opportunities to continue to work in their fie ld. Cuban Americans are the only Latina/o subgroup whose educational credentials were accepted and validated by the U.S. government, giving Cubans an advantage over other Latina/o subgroups to succeed in American society. In addition to federal assistance, the inner drive and agency of the Cubans helped them prosper as a community especially, in Miami, Florida. The different migration patterns and history among Latina/o communities is one of many examples of how diverse the Latina/o population really is. Ho wever, the intention of Chapter 2 is not to make a comparison between the three largest Latina/o population groups but to be cautious about the interpretations and generalizations of the not all Latina/os are first generation students or come from working class families. As previously mentioned, the college graduation rate for Cuban Americans is above the national average, surpassing White Americans (Census Bureau, 2011). The purpose of this study is to include all Latina/o groups. However, many of the studies referenced in Chapter 2 are based exclusively on a Mexican American sample. Unlike Puerto Ricans and Cubans, who are mainly concentrated in one or two specific cities and states, Me xican Americans who account for over half (65%) of the Latina/o population, are concentrated in multiple cities and states. For example, in a recent study, researchers noted that 68% of the Cuban population live in Florida and 23% of
45 Puerto Ricans reside i n New York (Ennis, Rios Vargas, & Albert, 2011). Therefore, it is are likely to include a large sample of Mexican American students. Thus, expanding the literature rev iew of Mexican Americans to be inclusive of other Latina/os is justifiable. For the remainder of Chapter 2 and for the purpose of this study, Latina/os will continue to be talked about as one group. In the 21 st century, the growth in the Latina/o populatio n was mainly due to the births of existing residents (60%) as opposed to Latina/o immigrants (40%) (Fry, 2008). Nonetheless, a large percentage of Latina/o students are very close to the immigration experience, even if they are not immigrants themselves, s ince many are the children or grandchildren of migrants (Pew Hispanic Center, 2009; Zwick & Sklar, 2005). Of those Latina/o students enrolled in college, 88% are U.S. born citizens (Santiago, 2007). Unfortunately, many Latina/o students lack the social and cultural capital to know about the value of college and the admissions process (McDonough, 1997) since many of their parents have less than a high school education. As a result, many researchers (Ganderton & Santos, 1995; Glick & White, 2004) have examin ed whether there is a relationship between the generational status of found that Latina/os who are born in the U.S. attend college at a much higher rate than those who we re immigrants (Rong & Grant, 1992; Feliciano, 2002). Ironically, some researchers have found that college attendance for Latina/o students does not increase with successive generations of U.S. residency (Hagy & Staniec, 2002; Rong & Grant, 1992). As common ly expected, Rong and Grant found Latina/o immigrants graduate
46 from college at a much lower rate compared to second generation Latina/os (U.S. born Latina/os with one or more immigrant parent). However, Rong and Grant did not find similar results among sec ond and third generation Latina/os. On the contrary, they found third generation Latina/os (U.S. born Latina/os with both parents born in the U.S.) graduate from college at a much lower rate than second generation Latina/os. Researchers should study whethe r a difference exists between third generation Latina/o students whose parents are college educated and those whose parents are not. Social and Cultural Capital include social and cultural capital (Ceja, 2006; Cohen, 2009). According to Bourdieu (1986), social capital is measured based on the quality of resources one has access to through their relationships with others or membership in a group. Bourdieu argued that of social capital one possesses greatly depends on the size of the network of those connections and the volume of the capital possessed by those whom one is the educationa l context, social capital refers to the ability for social networks to facilitate educational advancement for students (Bourdieu, 1986; Stanto Salazar, 2004) whether it be through their peers, educational agents, community members, friends, and/or family m embers. For example, students greatly benefit from the relationships they establish with the gatekeepers of their institutions, such as teachers, counselors, and other school personnel, because they can help them navigate the college choice process or dire ct them to other resources, since many are college graduates themselves. Institutional gatekeepers consist of anyone (school personnel) who can directly transit or negotiate the transmission of institutional resources and opportunities
47 towards students (St anto Salazar, 2004). It is through these social networks that Latina/o students are given the opportunity to acquire the necessary skills and (Saunders & Serna, 2004, p.148 ). When given the necessary social and cultural capital, Latina/o students are as likely as White students to enroll in college (L. Perna, 2000). According to Rios Aguilar and Deil cessfully through college if they either reconfigure their find ways students can create new relationships with college personnel. A growing body of research indicates that school counselors serve as social capital for high school students in terms of providing them with information and services that will help students apply to and be admitted to college. This is specifically true for Black and Latina/o students (McDonough, 2004; Bryan, Moore Thomas, Day Vines, & Holcomb McCoy, 2011). Institutional gatekeepers provided more than just resources to students, they provided valuable inf ormation, norms, and support (Coleman, 1988). To other students, educational gatekeepers serve as a social resource who can provide them with technical, psychological, cognitive, and informational assistance (Bryan et al., 2011). Thus, when it comes to obt provide the only source of social capital for low income students and students of color who are first Rios Aguilar and Deil Amen (2012) also found that Latina/o students were strongly encouraged by their social networks to attend college. However, upon
48 enrollment in college, many of their old social networks provided little guidance in planning for professional, career, and postgraduate option s after college. This is why it is critical for postsecondary institutions to offer Latina/o students more opportunities to establish new relationships with institutional agents that can help them transition into college, and afterwards into the workforce college enrollment rates may be attributed to lower levels of access to the types of suggested that a primary function of s ocial capital is to enable students to gain access to other forms of capital, such as cultural capital and institutional resources and support. Moreover, cultural capital refers to general characteristics, such as skills, knowledge, and traits that tend t status (L. Perna, 2006). One form of social capital that may be helpful to promote college enrollment consists of those students who have information in their homes about college (L. Perna, 2006). Simila rly, students who have knowledge about higher education possess some form of cultural capital which is considered necessary for suggested that the college decisions of Latina/o st udents are limited because many lack the type of social and cultural capital valued or needed in the college choice process (L. Perna & Titus, 2005; L. Perna, 2000). Thus, if members of the upper class tend to possess the most valued kinds of cultural and social capital (Bourdieu & Passeron, 1990; McDonough, 1997), then SES must be taken into consideration when discussing these particular issues related to Latina/o students. SES is commonly measured by parental income (Hearn, 1991) or includes parental inco me and education (A.F. Cabrera
49 2005b, 2005c) three part series on Latina/o students in the educational pipeline, the researchers found family income to be a major predictor of educational attainment postsecondary degree completion (Swail et al., 2005). Ganderton and Santos (1995) nding college. Latina/os may be at a great disadvantage. In general, Latina/o students come from families that are financially underprivileged when compared to the general po pulation. According to the Census Bureau (2012), the average annual earnings of Latina/os households is $38,624, well below the average earnings for all Americans of $50,054 in 2011. The highest level of education for the average Latina/o, age 25 and older is a high school diploma (62.9%), while the highest level of education for the average American is some college (Census Bureau, 2011). These findings highly suggest that the low socioeconomic status of Latina/os may have something to do with their low ed ucational attainment levels. English Proficiency postsecondary participation rates is that for a majority of Latina/o students, they live in a household where they, or in particular th eir parents, struggle with the English language. lack of English proficiency impacts their college choice process. First, some researchers have argued that Latina/o stude nts who do not speak English fluently may not perform well on standardized tests (Altshuler & Schmautz, 2006; Pennock Roman, 1990). Other
50 discourage some Latina/o students from contin uing their education beyond high school (Zarate & Gallimore, 2005). A third explanation is that Latina/o parents who speak limited or no English at home may be unable to assist their children through the college choice process and be unable to communicate effectively with high school or college personnel about college requirements and financial aid information (Ceja, 2006, Hurtado Ortiz & Gauvain, 2007). However, other researchers argue that Latina/o ason for the low college enrollment rates of Hispanic students (Pew Hispanic Center, 2010). After all, 88% of U.S. born adult children report that they speak English fluently ( Hakimzadeh & Cohn Introduction to College Choice Process Throughout the college choice literat ure, researchers have used various theoretical frameworks and models to provide insight on the stages students go through when making decisions about whether and where to enroll in college (Bergerson, 2009a) along with the factors that influence these deci sions. Some frameworks and attend college (Kotler & Fox, 1985), and many others have d escribed the entire college choice process (Hossler & Gallagher, 1987; L. Perna, 2006). The first half of this section starts off with a brief overview of the early theoretical frameworks used for the college choice process, throughout the literature. The second portion looks at the development of models that attempt to explain the college choice process, along with studies that used such models.
51 Early Theoretical Frameworks The descriptions of the college choice process have been based on three theoretical frameworks: psychological, sociological, and economic (Paulsen, 1990). The psychological framework focused on how environmental, institutional, and student characteristics of a college environment impacts student enrollment (Paulsen, 1990). More specifica lly, environmental characteristics included the population of college students, along with employment opportunities for college graduates versus non college cost of p ostsecondary institutions and how this influences what type of students get admitted to which institutions based on these characteristics. As for the component dealing with student characteristics, this includes parental education and income (Paulsen, 1990 ). The last two characteristics (i.e., institutional and student characteristics) will be f urther discussed, especially how they apply to this study. Whereas the psychological framework focuses on the institutional component, the sociological point of view focuses on the student. The sociological perspective personal characteristics. These characteristics included scholastic aptitude (Chapman, 1981), socioeconomic status (Chapman, 1981; Hearn, 1988) and parental income and education (Litten, 1982). Research studies that use the sociological theoretical their background characteristics (Ho ssler & Stage, 1992; Kao & Tienda, 1998). The sociological framework is salient to this study, especially as it pertains to Latina/o
52 Similar to the sociological framework, the e conomic viewpoint focuses on the student. However, the economic perspective views college choice as an investment like decision process, where students take into account the perceived monetary benefits and cost of college attendance (Paulsen, 1990). Throug h the economic framework, students decide to invest in college if the expected benefits outweigh the costs (HSB) used a sophomore cohort to uncover how tuition and financial aid affects assistance promoted enrollment among students, particularly grant aid. Unlike the two other frameworks, the economic perspective is the least talked about in t his study, like decision process were included in the national database used for this study. Development of College Choice Models These three theoretical frameworks provided the core founda tion for the development of various conceptual models that have helped to study the college choice tatus, academic aptitude and performance, and educational aspirations. The research also identified external influences, such as college characteristics, the influence of significant to communicate with students, as being influential in the decision by students about which college to for setting recruitment policy to identify the pressures and influenc es they need to
53 levels, and geographic location. Litten examined the different ways students approached and participated in the college choice process and found differences in the timing of the process. For instance, he identified how parental education influenced student selection process, along with the various ways college information was obtained. Litten affected their college selection process. Moreover, the majority of studies h ave used the college choice process known as the three phrase model (Hossler & Gallagher, 1987; Bergerson, 2009a). Based on a simplification of previous work, the model summarized the college choice process into three stages: (1) predisposition consists of the decision search for general information about college and learning about specific institutions; and (3) choice consists of students completing college applications and selecting a specific institution to attend (A.F. Cabrera & La Nasa, 2000a; Hossler et al, 1999). research to examine the college choice process of students from various backgr ounds. For example, Hurtado, Inkelas, Briggs, and Rhee (1997) used concepts from Hossler ethnic groups. Before the 1990s, most of the research on the college choice proc ess was based on the misleading assumption that all groups had equal access to college. However, more researchers have examined the significant differences in educational attainment across racial/ethnic and socioeconomic status (Bergerson, 2009a). Building
54 (1992) developed a model that outlined factors influencing each of Hossler and the model b y Hossler and Gallagher includes that of A.F. Cabrer and La Nasa (2000b). This model incorporated a phase known as when (i.e., which grade level) the college choice stages took place. By taking into consideration grade level, influential factors, and outco mes, both new models have demonstrated the flexibility of Hossler and students at each stage of their college choice process. Up to the 1990s, the college choice research was mainly shaped by three perspectives: sociological, psychological, and economic (Paulsen, 1990). The psychological perspective emphasized the environment of an institution, whereas the sociological approach examined college aspiration as part of a general attainment status. And finally, the economic perspective viewed college choice as an investment like decision process. The 1980s, on the other hand, saw a number of conceptual models being developed for studying the college choice process of students, in p the c omplexity of the college choice process for students of diverse backgrounds,
55 A more recent model, propos ed by Perna (2006), integrates aspects of economic model makes the assumption that student enrollment decisions are connected to their e are numerous paths that may lead to college enrollment. The proposed conceptual model (see Figure 2 1) is centered on the economic perspective which argues that college choice decisions are based on the evaluation between the expected benefits and the ex pected cost of college. Unlike model (2006) are not stages but rather interactive throughout the entire college choice a school and community context; (3) a higher education context; and (4) a social, economic, and policy context, which consist of the sociological perspective. Utilizing the archers to investigate various constructs related to the college choice process which come from both the sociological and economical models. used in this study due to inclusive ness of the diverse college student age population. Perna noted that many of the earlier models relied exclusively on quantitative data and often excluded the introspective data that typically emerges from qualitative data (Bergerson, 2009; L. Perna, 2006) Perna incorporated both qualitative and quantitative studies when developing her proposed conceptual model. Her model takes into college choice process at different rate s. This differs from earlier models that suggest all
56 students pass through certain stages at specific times (Bergerson, 2009). In spite of not being a linear model like Hossler and Gallagher (1987), Perna still suggests that possible outcomes such as savin g for college and academically preparing oneself are model is relatively young, it has already started to impact how researchers study the college choice process and al lows for various factors to be addressed all at once, whereas previously, they could only be studied in isolated situations (Engberg & Wolniak, 2009; Steinberg, Piraino, & Haverman, 2009). Layer One The first layer, individual habitus, consists of student beliefs, and perceptions that develop from the environment a student is most associated with. Collectively, these factors can influence the aspirations and attitudes that students have about postsecondary institutions (Bourdie conceptual model (2006) incorporates the influence of demographic characteristics, as well as social and cult ural capital (L. Perna, 2006). Demographic characteristics consist of gender, race/ethnicity, and socioeconomic status, all of which are pertinent to this study. Two additional constructs included in layer one are supply of resources and expected costs. Layer Two School and community context is the second layer which recognizes ways in which social structures and resources assist or thwart students during the college choice process. Past research has found that many students do not have the opportunity to attend high schools that have the necessary resources to successfully pursue college options (Engberg & Wolniak 2010), which stresses the need to further
57 investigate how the school and community context can influence the college choice process of Latina/o students. Latina/o students are more likely to attend public high schools in urban areas that are racially seg regated, have poorly maintained facilities, and have high rates of undertrained and under credentialed teachers (Aguirre & adequately prepare them for college, especially selective institutions. Many other students are overrepresented in several at risk academic areas such as: having a C average GPA, low test scores, frequently changing schools, and/or being held back (Swail, Cabrera, & Lee, 2004). This often results in lo wer levels of college preparedness for Latina/o students and to an extent helps to explain the underrepr esentation of Latina/o students in selective four year institutions. Therefore, the purpose of this study tion and type (i.e., private vs. public), Layer Three postsecondary institutions have on the decis ion by students to attend college (L. Perna, 2006). For example, postsecondary institutions often provide students and their parents with information that may influence their decision to attend a specific institution. In particular, institutional character istics, such as school rankings, programs of study, enrollment size, and campus location, can determine whether or not parents or students are interested in obtaining more information about certain institutions. This information will allow students and par ents to determine whether their goals and values are in alignment with the particular institution (L. Perna, 2006). This study will focus on those
58 specific characteristics Latina/o students are more likely to report as important in their college choice pro cess and in their decision to enroll in college. Layer Four The final layer, the socio economic and policy context, looks at how social forces (e.g., demographic changes), economic conditions (e.g., unemployment rate) and public policies (e.g., establish ment of a new need college choice process. Earlier models often failed to address how policy implications, process and which instituti ons students were able to consider based on these implications (L. Perna, 2006). Unfortunately, the specific dimensions presented in this layer are beyond the scope of the current study, as well as the dataset utilized. Therefore, this study will not inclu Perna does separate her model into four specific layers, each layer indirectly and directly influences the context of the other layers. Although this study only considers three of the four layers sugge factors, it is still important to look at the variety of influences this model acknowledges in refined in a large n umber of studies (Bergerson, 2009), it is still one of the most relevant models used for the purpose of this study.
59 Fig ure 2 Springer by Copyright Clearance Center
60 Figure 2 High school and community context (layer 2) Availability of resources and type of resources High school type and urbanization Structural supports and barriers Habitus (layer 1) Demographic Characteristics Gender Race/ethnicity Cultural Capital Value of college attainment Social Capital Information about college Assistance with college Social, economic, and policy context (layer 4) No variables examined in this study Higher ed ucation context (layer 3) Institutional characteristics High school context (layer 2) Degree of urbanization High school type Habitus (layer 1) Demographics Characteristics Gender Family composition Number of Siblings Cultural Capital Language Generational Status P Social Capital Demand for higher education No variable examined Supply of resources Family income Number of schools applied to Expected benefits No variable examined Expected costs Type of college When enrolling in college College Enrollment
61 College Choice Process of Latina/o Students This literature review examines the college choice process of Latina/o students in light of two contr adicting trends. For one, there has been a rapid increase in the college enrollment rates of Latina/o students. However, they continue to be one of the least college educated groups in terms of bachelor degree recipients. Since Latina/o students are more l ikely to be first generation college students compared to other racial/ethnic groups (Santiago, 2007), this literature review also includes research on first generation students. The college choice process of Latina/o students will be further examined usin g theoretical frameworks used for this study. Predisposition Studies that focus on the predisposition stage look at the formation of educational aspirations and intentions (A.F. Cabrera & La Nasa, 2000a). The underlying stage. Several studies highlight the importance of parents in the formation of college aspirations for Latina/o students But in general, parents play a key role in the predisposition stage of all racial and ethnic groups (Hamrick & Stage, 2004; Rooney, pectations to be a strong indicator of Latina/o Bruns (1998), studied eighth graders and found that higher levels of parent student interactions in learning activities, involvement in their c extracurricular activities, led to an increased number of school years those eighth graders were expected to complete.
62 Moreover, Azmitia et al. (1994) contrasted the educational aspirations Mexican American p arents had for their children against educational aspirations of White parents. They found that Mexican American parents held higher educational aspirations for their children, and many wanted their children to continue their education beyond high school. Gomz (2003) found that parental monitorin g, support, and help with schoolwork were associated with Latina/o students increased academic motivation and higher educational aspirations. But despite the high expectations many Latina/o parents have for their children, they continue to lack the inform ation (i.e., financial aid or college application procedures) their children need to maintain and realize college aspirations. Since a large percentage of Latina/os are first generation college students (Torres & Hernandez, 2009), many are in need of addi tional information and socialization to help them get into college (McDonough, 2004). This need arises from the large concentration of Latina/o parents who are not college educated or familiar with the American educational system, thereby rendering them in capable of passing on the necessary information their child needs to enroll in college and to be successful in their educational endeavors (Gloria & Segura Herrera, 2004). Nonetheless, social support for many Latina/o students comes from other family membe rs such as, siblings, cousins, and distance relatives. According to Gloria and Rodriguez (2000), familial support reference to academic persistence. For example, many Latina/ o students are influenced
63 by family members who have been through the educational pipeline and have helped them to acquire the necessary skills needed to survive college, along with providing academic support and guidance (Hurtado & Carter, 1997). Other st udies have found amount of emotional support and encouragement they received from their parents (Hurtado et al., 1996). Overall, the majority of the literature on Latin members, such as siblings, often play a significant role in the predisposition stage of Latina/o students. According to the lit erature, older siblings who have attended or graduated from college may play a critical role in the development of their younger sharing information about college, an d providing encouragement (M.T. Hurtado, 1997; Ceja, 2006). For example, in Over the Ivory Tower, Gandara (1995) interviewed fifty Mexican Americans who overcame various forms of disadvantages to reach the highest level of educational attainment: a Ph.D., M.D., or J.D. degree. Gandara found that many of these high achievers credited their older siblings with playing a significant role in transmitting college expectations. Despite many studies that recognized the important role of siblings in the decision to attend college, there is still much research left unexamined that will help us better understand the specific ways college educated siblings influence the college aspirations of their siblings. Another common factor found in the literature is the influenc e of SES on the predisposition stage of Latina/os (Hamrick & Sage, 2004; Kao & Tienda, 1998). Hamrick
64 and Stage (2004) explored the predisposition of students who attended predominantly minority and low income schools. The authors found SES, as measured by but not that of White students; emphasizing the need to better understand the college choice process of different racial/ethnic groups. Butner et al. (2001) fou nd differences in 31 32). Other researchers who suggested similar arguments are Jun and Colyar (2002); they proposed a connection between social networks, cultural capital, and SES as a predictor of educational outcomes. The researchers argue that the social class standing of low SES families hinders their ability to have well developed social networks and cultural capital that promote higher education. The literature has also found high schools to be influential in the formation of college aspirations (Ceja, 2001; Meredith 2008). Regrettably for many Latina/o students, high schools often discourage student educational aspirations instead of promoting higher education. Researchers like Butner and associates (2001) have noted that minority women (i.e., Latinas and African Am ericans) often face lower expectations from high school counselors than males. To support similar claims, L.X. Perez (1999) highlights the fact that teachers and counselors can often form roadblocks for parents spirations. A growing body of research indicates that high school counselors have tremendous influence on the college plans of Latina/o students. Unfortunately, these students are least likely to have counselors with
65 less training or those responsible for too many other, non counseling related tasks to give them good college guidance (McDonough, 2005; Plank & Jordan, 2001). Nonetheless, first generation students who receive academic support from their school counselor, as well as information about college programs, and assistance with the college admissions process are provided with the social capital that can compensate for family networks when students' parents have limited resources or education (Bryce, Moore Thomas, Day Vines, & Holcomb McCoy, 2011). In this same study, using the ELS:02 data, Bryce, Moore Thomas, Day Vines, and Holcomb McCoy 10 th grade were more likely to apply to college compared with students who did no t see Hispanic students who did not go to their school counselor for college information were more likely to apply to two or more schools compared to Hispanic students who saw the counselor after 10 th grade (Bryce, Moore Thomas, Day Vines, and Holcomb McCoy, 2011, p.194). saw potential in them and encouraged them to attend college (Butner et al., 2001; Rooney, 2008). In recent years, teachers have received heightened attention in light of proposals to base their salary and promotion on measures of teacher quality and student performance (Loeb, Rouse, & Shorris, 2007; Kyriakides, 2005). Gin orio and encourage, and respect students, related to the academic achievement of Latina/o students. Thus, the role of teachers and counselors could be further studied in order to
66 understand the role they play in the predisposition stage of Latina/o students. Another area of study could include high school coaches, where a dearth of research exists. In addition to school factors, researchers have started to examine how social and c ultural capital affect Latina/o students in the predisposition stage. Some studies focused on the social relationships from which a student could potentially receive various types of resources and support (Ceja, 2001; Cohen, 2009; Stanton Salazar & Dornbus ch, 1995). Stanton aspiration of Mexican American high school students and found a connection between levels of social capital and school based ties. In particular, Stanton Salazar and Dornbusch (1995) fou nd that students with higher educational aspirations also had higher levels of social capital. For example, these students had more high status contact with adults who had access to college related information and support from teachers, counselors, and oth er school personnel among the networks of people they knew who could provide college related informational support. Thus, many students may develop college aspirations sooner, if this is transmitted as an expected behavior s. Perez and McDonough (2008) noted that social close shut the college going doors for Lati of how social networks, such as parents, siblings, peers, and high school personnel act as influential resources during the college choice process of Latina /o students More specifically, Perez and McDonough (2008) have reported peers to have both positive
67 have provided conflicting findings regarding the impact peers have o college aspirations (Contretras Godfrey, 2009; Gomez, 2005). In their examination of the influence of peers on 6 th and 7 th graders in a community college academic outreach program, Azmitia and Cooper (2001) found peers represented a ch allenge to college aspirations. Similarly, Gomez (2005) found peers negatively impact the college aspirations of Latina/os who did not plan to attend college. Conversely, study found that students were often encouraged by their friends to at tend college. Contreras concurred with As proven by various studies, peers matter in the predisposition stage of Latin a/o students, but whether they positively or negatively influence the values and behaviors of Latina/o students to attend college varies. The literature suggests that peers have a positive influence on college aspirations and attendance, but usually when s tudents are (Contrera Godfrey, 2009; DiMaggio & Mohr, 1985; Gomez, 2005). To an extent, cultural capital makes it possible for students to have access to such peer group part icipation with those who value higher education and frequently engage in conversations that is cultural cap ital. Cultural capital refers to the values, knowledge, skills, and abilities Unfortunately, many low income Latina/o students lack the cultural capital of middle and
68 upper cla ss students, often resulting in lower educational aspirations and college preparation rates (Jun & Coylar, 2002). Tierney (1999) supports this claim with an examination of a college preparation program that used several strategies to enhance ural capital. The program did this by stressing to low income students that T he program attempted to instill a form of cultural capital that low income students often lack in comparison to middle and upper class students who are regularly told by their famil ies and friends to further their education. Overall, the underlying principle of this research is to acknowledge that Latina/o students experience the college choice process differently from other racial/ethnic groups a fact which requires further investigation lack of information about financial aid or college application procedures makes it difficult for them to help their children fulfill their college aspirations (Azmitia et al., 19 94; L.X. college bound classes often hinders their college aspirations (Gonzalez et al., 2003; P.A. Perez, 2007) while do not plan to continue their education beyond high school (Azmitia & Cooper, 2001; Gomez, 2005). influence on their college aspirations especially, when it includes in dividuals who limit McDonough, 2008). Nevertheless, many Latina/o students develop and maintain their college aspirations which they carry on to the search stage. Sear ch The second stage of the college choice process looks at where Latina/o students gather information about college and whether they talk to parents, peers, school
69 counselors, teachers, coaches, relatives, and/or college admission recruiters as they consid er which college they plan to attend (Hossler et al., 1999). This section addresses the financial aid process, and their perceptions about being able to pay for college all of which are greatly influenced by the type of social and cultural capital studen ts have access to. In general, Latina/o students are usually uninformed or misinformed about college and financial aid (Admon, 2006; Kao & Tienda, 1998). Many first gene ration Latina/o students do not possess the social and cultural capital necessary to obtain facts about college and financial aid. Whereas, most students may rely on teachers, guidance counselors, and college admission personnel, as was found in a nine yea r study of high school students (Hossler et al., 1999), Latina/os may seek other sources (Admon, 2006; P.A. Perez, 2007), such as help from relatives or friends who may have some knowledge about college. A possible explanation about why Latina/o students are less likely to rely on school counselors for college information may have to do with the large school to counselor ratios at many high schools where there is a high Latina/o enrollment rate. Hence, counselors may not be readily accessible for one on on e counseling sessions with students. According to the American School Counselor Association (n.d.), in the four states where Latina/o students make up the largest share of the population (New Mexico, Texas, California, and Arizona), the recommended student to counselor ratio of 250 to 1 was exceeded by more than half, with Arizona having the highest student to counselor ratio of 815 to 1. Although school counselors are often reported to play a is may not be the case for
70 many Latina/o students. According to Ceja (2001) and Cohen (2009), many Latina/o students are seriously underserved by their counselors when it comes to obtaining college information or pursing a college education. Other studies have found that counselors often track Latina/o students into vocational paths as opposed to college bound tracks, which may deter some students from pursuing a college degree through their discouraging words or behavior which mirrors low expectations ( Vel a Gude et al., 2009 ). The negative relationships that generations of Latina/o students have seek college information from guidance counselors ( Vela Gude et al., 2009 ). Lat to afford college also influence the type of school a student is willing to consider and learn more about (A.F. Cabrera & La Nasa, 2000a). Unfortunately, most Latina/o studen ts and parents know less about the financial aid process and the cost of college than other racial/ethnic groups (Kao & Tienda, 1998). This lack of information Latina/o students and parents have about college cost and financial aid options often results in a distorted perception of their ability to afford college (L.X. Perez, 1999), in particular four cost institutions, work more and borrow less than students at other instituti 2005, p.5). According to Berkner et al. (2005), only 12 percent of community college students take out student loans versus 40 percent of students at other institutions. For many low income and first generation students, debt aversion freque ntly begins with parents and their reluctance to borrow as a result of their own negative experiences with
71 past indebtedness, their leeriness about giving out financial information, or their limited experience with credit (Burdman, 2005). In her study, Bu rdman (2005) found that among full time students, first generation students were twice as likely as those students whose parents had college degrees to work full time instead of borrowing. Financial literacy may help in properly educating first generation Latina/o families about borrowing student loans. While most high school students tend to rely on their high school counselors for financial aid information, many school counselors report that they do not feel comfortable with the intricacies of financial a id policies, especially loans (Burdman, 2005). Many other high school counselors are too busy serving hundreds of seniors at understaffed schools and find little time to address the issue (i.e., financial aid process or literacy) while many others do not h ave the proper training to do so (Burdman, 2005). differently from other students, with many first generation Latina/o students lacking information about the financial aid p rocess and the cost of college (Admon, 2006; Gomez, 2005). Nonetheless, many Latina/o students successfully navigate through the search stage of the college choice process. However, who exactly Latina/o students go to for college information is important t o examine in order to better understand how they navigate the search stage. Researchers have suggested that Latina/o students often sought parents, close family members, counselors, teachers, and peers for guidance ( Stanto Salazar, 2004; L. Perna, 2006). H college information is still left unexamined. Therefore, further research is recommended
72 to look into the role parents, close relatives, counselors, favorite teachers, and peers stage, as well as their choice of college. Choice The final stage is the choice stage, which consists of the ability of students to make decisions about where to apply and enroll in college (A.F. Cabrera & La Nasa, 2000a; Hossler & Gallagher, 1987). This section reviews the research that is related to parental factors, cost of tuition, financial aid, distance to home, familial obligations, and academic preparation influe education, and level of involvement with th eir students (M.T Hurtado, 1997; S. Hurtado educational attainment, and employment statu s) was highly associated with the type of institutions students considered a potential college choice. In particular, the authors submitted but their educational levels were found to have no significant influence. Similar studies found that SES is an indicator of college enrollment for Latina/o students found that students are more likely to apply to more colleges if they came from In their overview of research on racial, et hnic, and immigration differences in
73 educational achievement and attainment, Kao and Thompson (2003) argued that (p.431). The researchers noted that low income Latina/os ar e disproportionately represented in community colleges suggesting that Latina/o students are more likely to select community colleges as a result of being low income and not necessarily due to personal preferences. Additionally, parental involvement and ex pectations significantly influence the parental involvement related to college enrollment but it varied by race/ethnicity. The authors used the National Educational Lon gitudinal Study (NELS) and operationalized parental involvement employing two indicators: parent student involvement and parental monitoring. Overall, the odds for any student to enroll in college increased with t he frequency with which parents discussed e ducational related topics, volunteered at the school, and contacted the school about academics. Therefore, higher frequencies of year capital that provides individuals with access to resources that may facilitate college involvement may be explained by paren educational expectations for their children manifested in a significant effect on the
74 Furthermore, conventional examples of parental involvement are inadequate in (Lopez, 2001; Kiyana, 2008). According to Lopez (2001), Latina/o parents may expose their children to manual labor in order for them to witness the consequences of not having a college degree. This is intended to teach the value of education. In addition, older siblings assist younger siblings with homework assignments (Kiyama, 20 08). Lopez noted that if these behaviors (parental involvement) were viewed from a education. Instead, Lopez describes Latina/o parents who were involved in their children Other factors supported by the literature include the cost of tuition and financial aid (Carreras, 1998; Ceja, 2001). While students in general may overestimate the cost of college (Kirst & Venezia, 2004), financially disadvantaged students may be more likely to overestimate the costs of college attendance (Grodsky & Jones, 2007; L.X. Perez, 1999). In particular, Kirst and Venezia (2004) found that low income students were four to six times more likely to overestimate the cost of tuition. In their study, Ganderton and Santos (1995) found that increasing the capacity of Latina/o families to finance college increased the probability that more Latina/o students would enroll in and Santos and suggested that in order to increase persistence and college graduation, more efforts should be made to increase financial aid for Latina/o students. Kurlaender (2006) found that limited financial resources may be one reason Latina/os select
75 community colleges over four year institutions, especially since many are financially disadvantaged and thu enrollment and found that first generation Latina/o students tended to enroll in the attend college, and f inancial aid was found to have a significant influence on their decision to enroll in college. Geographic location is another important factor in the college choice stage year colle ges that first generation students are more likely to attend are based primarily on cost of attendance and distance from home. In this study, gender differences arose, with Latinas being more likely to attend a college closer to home whereas Latino males w study of gender differences in regards to geographic location on college choice, with Latinas mentioning proximity to home more often than Latino males as a very important factor towards their college choice stage. emphasis on keeping close proximity to home may explain why many choose community colleges over four year institutions (Kurlaender, 2006). Many other Latina/o students may choose to attend comm unity colleges due to family obligations and not necessarily because of the cost of college (Ceja, 2001). Parental expectations may be another reason many Latina/o students decided to attend
76 a college with close proximity to their homes (Lopez Turley, 2006 ). In her study, Lopez education: college at home or college anywhere. Lopez Turley found that Latina/o parents are more likely to be college at home parents. Parents lack of familiarity with year institutions over two year institutions may be another reason why Latina/o parents are more hesitant to allow their children to go off to college. Fina application and enrollment behavior decisions (L. Perna, 2000; Swail et al., 2005c). For example, Hurtado and associates (1997) found that Latina/o students in vocational programs sub mitted fewer college applications, and yet were more likely to apply to for profit and community colleges in comparison to Latina/o students who were in rigorous academic abi college enrollment rates. The study even found that students who were enrolled at four year colleges exhibited higher academic ability in comparison to students who were not in col lege. For similar reasons, Kurlaender (2006) argued that weak academic preparation may be a reason some Latina/o students choose to attend community academic preparation wa s left unmeasured due to the complexity of trying to measure for this factor. differently from other students. Researchers have found that parental education,
77 income, and involv ement, along with college cost, financial aid, geographic location, choosing a college to attend. Although the literature suggests that some of these factors positively in present. Nevertheless, many Latina/o students successfully complete the college choice process by enrolling in college. Where students decide to enroll and whether they enroll part time or full time are equally important to know because this helps to determine Summary of Theoretical Frameworks This study incorporates elements from two college choice models: Hossler & conceptual model of studying college choice. The two models assist in the development of the literature review, as well as the structure of the research questions and aid in the analyses of model has framed much of the college choice research to date and because it simplifies the college choice process into three stages: predisposition, search, and choice (Hossler et al ., 1999). In addition, this model looks at the college choice process from characteristics affects their college choice process. However, like any other model, there is always critic ism. For this model, it was the assumption that all students have equal access to college information and resources, failing to fully explain the experiences of all students, especially students of color, first generation students, low income students, and at risk youth (Bergerson, 2009b; A.F. Cabrera & La Nasa, 2001; Talavera Bustillos, 1998).
78 does not have equitable access to the information and resources deemed necessary to success model proposes four layered contexts which influence the college choice process. However, for this study only three of the four layers of the model will be incorporated; individua social and cultural capital variables from the habitus layer, the school context from the school and community layer of the model, and the institutional characteristics of the hi gher education context were the most applicable to use in this study. framework used for this study with inc lusion model. As previously me process and that were made available in ELS:02 were used for this study. By implementing various variable social, educational, and personal variables that impact the college choice process of Latina/o students.
79 CHAPTER 3 METHODOLOGY The purpose of this section is to describe the research methodology util ized in this study to investigate the role social networks have on the college choice process of Latina/o students. C hapter 3 revisits the purpose and research questions guiding this study. The following section describes the source and sample. The depende nt and independent variables are further described, along with the statistical methods used to analyze the variables. Additional methodological considerations that have important implications for the research design of the study and data analysis are furth er explored. C hapter 3 ends by addressing the limitations of th is study. The purpose of this study was to examine the influence social networks have on s given to how the expectations of soc ial networks as determined by the student, influence d the predisposition, search, and choice stage a long with their decision to enroll in college. More specifically, each student was asked what they think their mother, father, friend, close rel ative, school counselor, favorite teacher, and coach desire d for them to do after high school. Students were asked this question twice, once as sophomore s and the second time as seniors. For the purpose of this study, it is fundamental to keep in mind that these desires or expectations are based on the verbal or nonverbal message s students receive d from the following social networks listed above. Moreover, t here were four research questions guiding this study: how the expectations that social networks have for Latina/o stude nts influence the ir college aspiration s ; how the expectations s ocial networks have for Latina/o students influ ence who students go to for college in formation; how the expectations social networks have
80 for Latina/o students influence the main reason students select ed a particular college; and finally, how the expectations social networks have for Latina/o students influence d students decision to enroll in college a fter high school. Proposed Hypotheses Based upon the research questions gui ding this study and the literature on Does the expectations social networks have for Latina/o high school students, along ee aspirations? Does the expectations social networks have for Latina/o high school students, along with other key variables, influence who their more likely to go to for college entrance information in both the 10 th and 12 th grade? Does the expectations social networks have for Latina/o high school students, along with other key variables, influence the primary reason students selected their chosen college? Does t he expectations social networks have for Latina/o high school students, along with other key variables, influence their decision to enroll in college? Data Source The data analyzed for this study comes from the National Center for Educational Statistics (NCES), Educational Longitudinal Study of 2002 (ELS:02). When this study was conducted, ELS:0 2 offered the most robust and recent national level data relevant Longitudinal Study of the High School Class of 1972 (NLS:72), the NCES began longitudinal data that linked th e educational experiences of high school students with later outcomes such as workforce experiences and collegiate experiences. The intention of these studies was to gather information on potential factors, such as family, cultural, economic, and social t hat may influence the development of students at different stages. ELS:02 provided longitudinal data for participants in three waves: the
81 the second follow up was in 2006 (students in postsecondary school or the workforce). The survey was administered to students, p arents, teachers, and administrators (principle and library media center director) at the student s respective high school. Howe ver, for the purpose of this study only Latina/o students were of interest The intention of ELS:02 was to monitor the transition of a national sample of high school students as they progress from tenth grade through high school and on to college and/or th e workforce and the social factors that influence the m I selected the ELS:02 dataset as the data source because it focused on the college choice process of d investigated the influences s ocial networks have college choice process: predisposition, search, and choice. ELS:02 also utilized a nationally representative sample allowing for the study to address the research questions in a timely and cost effective manner. Tenth grade students attending different types of high schools were included for ELS:02 These high schools included public, private, and Catholic schools with in the 50 states along with the District of Columbia. Of the 1,268 sampled schools, 1,221 were eligible but only 752 of these schools responded with a 67.8% response rate. The majority of schools were public high schools (88%) located in urban cities (53%). The t argeted population include s 10 th grade students who attende d study eligible schools in 2002 spring t erm Therefore, 752 samples schools across the United States gave permission for the administration of ELS:02, with 16,197 students participating among all these schools.
82 Sampling Design The majority of national datasets use a complex survey design in ord er to employ multistage, cluster, and/or stratified sampling strategies (Hahs Vaughn, 2007). Small scale empirical research studies typically use simple random sampling to ensure each subject in the population has an equal chance of being included in the s ample. Complex sampling designs, used by most national datasets, usually oversample subjects and institutions of interest therefore, they will have a higher probability of selection. Oversampling ensures subjects and institutions of interest are included i n the sample with sufficient numbers for the purposes of statistical analysis (Thomas & Heck, 2001). For example, non public schools specifically Catholic and other private schools, were sampled at a much higher rate than public schools in order to ensu re the sample wa s large enough to support comparisons between schools. Asian and Hispanic students were also sampled at a much higher rate than White and Black students in order to ensure a large enough sample for comparison among the se groups. A stratified systematic sample was used to select students on a flow base as student lists were received the strata included Hispanic, Asian, Black, and o ther race/ethnicity. The sample for ELS:02 was established using a two stage sampling design (Wei et al 2009). The first stage identified the institutions that would be included in the sample from approximately 752 eligible United States high schools. The institutional sampling frame was constructed with the intent ion to match the targeted population. The second st age of the sampling design consisted of selecting students from those 752 institutions with approximately 26 students per school being selected from these lists. Sample instructions provided students enrollment list s that were used to construct the samplin g frame. Approximately, 16,197 students participated in the ELS:02 base
83 year. Further information about the sampling design is available in the ELS:02 base Data Sample For the purpose of this study only Latina/o students who i dentified as Hispanic and were enrolled as tenth grade students at eligible schools during the 2001 2002 school ye ar (n=2209) were included in this sample. For this study, it was determined that first neither mother nor father had earned a college degree (Ishitani, 2003). Additionally, I felt this was the clearest way to discern students who may have gained advantages from parents who because I was interested in studying Latina/o students and their within group differences (first generation, socioeconomic status, social capital) Therefore, Latina/o students who did not classify as first generation college students were not omitted fro m the sampling frame but rather used for comparison purposes with first generation Latina/o students. However, conducting statistical analyses using such a large number of cases poses great risk for committing Type I errors (Thomas & Heck, 2001). Using a n ormalized or relative weight when analyzing large, secondary datasets is a common practice among researchers and helps correct for oversampling in the survey design (Hahs Vaughn, 2005; Thomas & Heck, 2001). The normalized weight reduces the overall sample size for statistical purposes, but still preserves the appropriate proportions of the complex survey design. In this study, a normalized weight will be used to conduct all statistical analysis.
84 Descriptive Statist i c s The following tables provide d descript ive information about the demographic variables of the sample of interest. As depicted in T able 3 1, there is a fairly equal representation of Hispanic male students (49.5%) enrolled in t he 10 th grade during the 2001 2002 school year in comparison to fem ale students (50.7%) Another key demographic variable for this study was t he estimated income level of Hispanic famil ies during the year 2002. ELS:02 originally provided 13 different income groups but I recoded this variable to included only four differen t income groups based on the distribution of frequencies. The largest group (51%) indicated that the total family income during 2002 was less than $35,000. The last two groups represented slightly higher income groups with 14.5% indicating a total family i ncome between $50,001 and $75,000 and 15.4% indicating a family income of $75,001 or more. Income groups are reflected in Table 3 2 and display a trend which indicates that higher income groups have fewer first generation students. Table 3 1 Frequency of gender (n=2209) Gender Frequency Percentage Male 1117 51% Female 1092 49% Table 3 2 Frequency of family income levels (n=2209) Income range Frequen c y Percentage Less than or equal to $35,000 1128 51% $35,001 $50,000 420 19% $50,001 $75,000 320 14.5% $75,001 or Greater 341 15.4% Table 3 3 displays further information about the family composition (BYFCOMP) of Latina/o students. ELS:02 originally provided nine categories representing different family compositions (BYFCOMP) but I recoded th is variable to include only six
85 categories: both parents (reference), one parent and one guardian or two guardians, mother only, father only, only one guardian, and parent or guardian lives with student less than half the time. The majority of Hispanic stude nts (54.91%) lived with both their parents, followed by 20.14% who lived with at least one parent and guardian or with two guardians, and le ss than 18.92% lived only with t heir mother. Further information about ructure is displa yed on T able 3 4 which includes the number of siblings Latina/o students have T here was no differentiation between half siblings or step siblings. ELS:02 originally provided eight options for the number of siblings (BYSIBHOM) Latina/o students have but th ey were recoded to include only four categories Most Latina/o student s (47.62%) report to have one to two siblings, 16.11% had three to four siblings, a similar amount (13.12%) had no siblings, and the lowest percentage (2.9%) had five or more siblings. T able 3 3 Frequency of family composition (n=2209) Family Composition Frequenc y Percentage Both parents 1213 54.91% One parent + Guardian or Two Guardians 445 20.14% Mother only 418 18.92% Father only 71 3.2% One Guardian 38 1.7% Parent/ Guardian lives with student less than half the time 24 1.1% Table 3 4 Frequency of number of in home siblings (n=2209) Number of in home siblings Frequenc y Percentage Zero 290 13.12% One Two 1052 47.62% Three Four 356 16.11% Five or more 66 2.9% Tab le 3 5 displays ad ditional information on student s native language. ELS:02 originally pr ovided six possible native languages : English, Spanish, other European language, West/South Asian language, Pacific Asian/Southeast Asian language, or
86 other language However, I recoded this variable (BYHOMLNG) to include only the top three most common native languages spoken by Latina/o students: English, Spanish, and other. The majority of their native language w as English (47.12%) followed by Spanish (46%), and the smallest percentage (2.35%) consisted of another language. Table 3 provided five possible categories that parents may fall under: not fluent English speakers partially fluent, fluent, no n native English speaker ( fluency unknown ) o r native English speaker. Paradoxically, the majority of Latina/o parents (37.1 2%) were native English speakers T he se cond largest group consisted of parents (27.79%) who were not fluent in Engl ish. The thir d highest group (1 6.65%) reported to be fluent English speakers According to these descriptive statistics, over half (53.77%) of all Latina/o parents reported to speak English fluently, whether they were native English speakers or not. Table 3 5 Frequenc Frequen c y Percentage English 1041 47.12% Spanish 1017 46% Other 52 2.35% Table 3 6 Frequency of parents English fluency (n=2209) Frequenc y Percentag e Fluency unknown 44 1.99% Not fluent 614 27.79% Partially fluent 50 2.23% Fluent 368 16.65% Native English Speaker 820 37.12% Table 3 7 provides information about the generational status of Latina/o students. ELS:02 variable (BYGNSTAT) include s thr generational status: (1) newly immigrate s (i.e. the sample member was born in Puerto Rico or a non US country, regardless of their mother's birthplace); (2) first generation
87 (i.e. the sample member was born in the United States, but his/her mother was born in Puerto Rico or a non US country); or, (3) second or more generation s (i.e. both the sample member and their mother were born in the US). The majority (35.8%) of Latina/o students reported to be second gen eration 26.7% consisted of students that reported to be first gener ation, and only 22.4% reported to be newly immigrates Table 3 7 Generational Status Frequenc y Percentage Newly immigrated 495 2 2.40% First Generation 590 26.7% Second Generation 791 35.8% Table 3 8 displays the different types of high schools th e sample population a ttended. The majority of students were enrolled in public schools (85.4%) followed by private schools (11%), and a smaller percentage of Latina/o students were enrolled in private Catholic schools (3.5%). The is reflected o n T able 3 9 The majority of Latina/o students attended schools in urban cities (47%), followed by suburban cities (44.3%), and the smalle st percent resid e in rural areas (8.6%). Table 3 8 Frequency of type of school (n=2209) Type of School Frequenc y Percentage Catholic 244 11% Other private 78 3.5% Public 1887 85.4% Table 3 9 Frequency of school urbanicity (n=2209) Location Frequency Percentage Rural 190 8.6% Suburban 979 44.3% Urban 1040 47% Table 3 1 0 displays information about the educational obtainment of Latina/o parents. ELS:02 originally provided 8 different educational levels : did not finis h high school, graduated from high school or GED, attended 2 yea r school with no degree,
88 graduated from two year school, attended college but no four year degree, graduated or other advanced degree. However, I recoded this variable (BYPARED) to included only five different educational levels based on the distribution of frequencies. The largest group (42.6%) of Latina/o parents reported to have a high school diploma or less as the hi ghest level of education obtained. The second largest group (21%) included parents Latina/o parents who held some type of advanced degree (i.e., Phd, MD, JD). Table 3 1 0 Freq First generation Frequenc y Percentage Advanced Degree 230 10.4% 320 14.5% 465 21% Some College no degree 252 11.4% High school diploma or less 942 42.6% Additional information concerning whether students plan to continue their education right after high school or postpone it ( T able 3 1 1 ) ELS:02 asked students, with five possible choices to select from : yes, right after high school; yes, after being out of high school for 1 year ; yes, after being out of high school over 1 y ea Over half of all L atina/o students (58.57%) said they plan to attend college immediately after high school. The second largest group (13.8%) consisted of students who plan to attend college one year after high school. Fewer students (2.1%) reported that they plan to attend college over one year after being out of high school Only 177 or 3.5% Latina/o students reported that they did no t know when they plan ned to attend college
89 Overall, the majority of Latina/o students (78.47%) plan ned to attend colle ge sometime in the future. T able 3 11 illustrates that the longer students wait ed to attend college the smaller the percenta ge who anticipate d a ttending college at all. Table 3 1 1 Frequency Frequenc y Percentage After high school 1294 58.57% One year later 305 13.8% Over one year later 47 2.1% 177 3.5% 12 0.5% Table 3 1 2 ned to attend. This variable was not recoded from its original categories: two year college s four year college s and vocational technical trade school. Over half (59%) of Lat ina/o stud ents planned to attend vocational technical trade school, followed by (12.22%) who plan ned to attend a two year institution and the smallest percentage (5%) of Latina/o students plan ned to attend a four year college/university. Table 3 1 2 Frequency of t ype of college to attend (n=2209) Type of College Frequenc y Percentage Four year college or university 112 5% Two year community college 270 12.22% Vocational, Technical, or Trade school 1305 59% Table 3 1 3 obtains information on the number of c olleges Latina/o students applied to. This variable was not recoded and the original 11 variables remain the same H owever, for demonstration purposes T able 3 13 includes a summary of the variables. The majority (22.67%) of students applied to at least one school, 15.75% of Latina/o students app lied to two schools, and 9.14% of students applied to three schools. Less than 20% of students applied to more than four schools.
90 Table 3 1 3 Frequency of schools applied to (n=2209) Schools applied to Frequen c y Percentage One 501 22.67% Two 348 15.75% Three 202 9.14% Four 106 4.79% More than five 189 8.5% Since I am only interested in Latina/o students, students who are not Latina/os This r educ ed the sample fro m over 16,000 to approximately 2,209 students. Nonetheless, this is still an adequate sample size that consists of a national database wh ich can be generalized to Latina/o students across the country. According to the descriptive statist ics, 42.6% of Latin a/o s in this sample fell under the category of first generation college student Contrary to the literature, there were more Latina/o students who had college educated parents (57.4%). For the purpose of this study it was determined that first generation students would be defined as students who neither parent earned a college education which was similar to previous studies (Ishitani, 2003). Dependent Variables The dependent variables used in this study to examine the college choice process of Latina/o s tudents are: (a) Latina/o students aspiration to attend college or not, (b) who Latina/o students went to for college information, (c) the ma in reason they chose their given college, and (d) whether they enrolled in college or not. Table 3 1 4 provides a su (1987) college choice model.
91 Table 3 1 4 Summary of dependent variables ELS:02 Source Variable Variable Type Scale College aspiration to attend college or not BYS56 Ordinal /Categ orical Multilevel analysis Recoded: 1=high school or less, 2= Associate degree, 3= attended four year, incomplete, 5=Advanced degree W ho 10 th grade students go to for information about the entrance requirements of various college s BYS59 Binary, Categorical Multilevel logistic regression Recoded: 0=No, 1=Yes Recoded Categorical Option: (a) Guidance counselor, (b) favorite teacher/coach (c) parents, (d) siblings other relatives, (e) friends, (f) college representatives, (g) libra ry, (h) none of the above. Who 12 th grade students go to for information about the entrance requirements of various colleges F1S48 Binary, Categorical Multilevel logistic regression Recoded: 0=No, 1=Yes Recoded Categorical Option: (a) Guidance counselor (b) Teacher (d ) parents, ( e ) Family (g ) friends, ( h ) C ollege representatives, ( k ) L ibrary, (n ) none of the above. Main reason students decided to attend their chosen college F2B14 Multinomial/ Categorical Categorical: (a) Program of study (b) reputatio n of school (c) cost (d) location (e) personal or family reasons (f) another reason. Enrolled in college F2B07 Dichotomous Binary:0= No, 1= Yes The first dependent variable focused on whether Latina/o students had aspirations to attend college. The va riable BYS56 asked students how far in school they th ink they will get and students answer e d with (a) less than high school graduation, (b) high school graduation or GED only, (c) attend or complete 2 year college/school, (d) attend college, 4 year degree degree or equivalent, (g) obtain PhD, MD, or other advanced degree. However, I recoded this variable into a multinomial categorical variable that divided the responses into f ive areas based on the type of degree students were most likely to aspire to (a) ttend four year but incomplete, (d)
92 dvanced degree. This dependent variable was used to examine the factors associated wi th student 1 5 provides frequencies for this variable A pproximately third fourths (74%) of Latina/o students in 10 th grade aspired to some type of college education. However, the majority of Latina/o students (33.1%) aspired to obtain a bachel who aspired for an advanced degree. The dependent variable was used to answer the first research ques tion presented earlier in C hapter 3 concerning the college choice process of high school students. Tabl e 3 1 5 s in 2002 (n =2209) Aspiration college a spirations Frequency Percent Aspire for high school or less 223 10.1% 118 5.3% Aspire for a Bachelor 125 5.6% 731 33.1% Aspire for an Advanced degree 657 29.7% The second dependent variable focused on the second stage of the college choice process and takes into account the influence of s ocial networks and who students are more likely to seek for college information. The variable named F1S48 asked students where they have gone for information about the entrance requirements of various colleges S tudents answers were binary (yes or no); (a) guidance counselor, (b) teacher, (c) coach (d) parent, (e) brother or sister, (f) other relative, (g) friend, (h) publications, or website, (k) school library, ( l) public library, (m) college or university library, (n) none of the above. However, I recoded this variable into seven binary options to reflect the interest of this study along with the support of the literature: (a) guidance
93 counselor, (b) teacher/coac h (c) parent, (d) siblings/other relative, (e) friend, (f) college representatives, (g) library, (h) none of the above. T he student sample used to examine the second research question was delimited to only those high school students who reported that they intended to continue their education after high school by answering question FIS45; Do you plan to continue your education at some time in the future? Students who responded no or did not know skipped question F1S48. Therefore, only Latina/o students who answered yes to continuing their education sometime in the future answered question F1S48 (n= 1204). Table 3 1 6 provides descriptive information on the top three sources Latina/o students seek for college information. Approximately three fourths (76.7%) of students went to their guidance counselor for college information. This was followed by teachers (56.9%) and college representatives (55.6%), both who shared very close numbers in terms of the most frequent sources 10 th grade students went to for college information. Table 3 1 6 Frequency of college search information (n= 1204) Top three sources College information Frequency Percent Guidance Counselor 924 76.7% Teacher 686 56.9% College Representatives 670 55.6% The third dependent variable focus off of F2B14 asked s tudents which of the following categorical options is the main reason they decided to attend their chosen college: (a) program of stu dy (b) reputation of school (c) cost (d) location (e) personal or family reasons (f) another reason. This variable was not recoded but for demonstration purpose s T able 3 1 7 only provides information on the three main reasons students selected their primary institution. The majority of Latina/o
94 students (14.6%) selected location as the main reason why they choose thei r c ollege, program of study was second (11.8%), and cost/affordability (10.7%) was ranked third. All three of these characteristics are support ed by the literature as reasons to why Latina/o students select ed the college they did, primarily community colleges. These type of institutions were chosen b ased on the ir close proximity to home and the low tuition cost in comparison to four year institut ions. Unfortunately, the type of institution students actually enrolled in was not included Table 3 1 7 Frequency of college characteristics Top three reasons College Characteristics Frequency Percent Location 324 14.6% Program of Study 261 11.8% Cost/affordability/financial reasons 238 10.7% The final dependent variable focused on whether Latina/o students successfully navigated the college choice process and actually enrolled in college. Variable F2B07 was used because it asked students had they ever attended a postsecondary school after high school. The binary question was not recorded for research purposes. Table 3 1 8 provides information about the percent age of students who actually enrolled in college Of all the Latina/o students in this sam ple, approximately 53% enrolled in some type of college following their high school graduation. Students were not asked what type of institution they enrolled in, such as public, private, four year, vocational/trade, liberal arts, or community col lege Moreover, Latina/o students who did not enrolled in college made up 30% of the overall sample population. Conversely, whether this sample of students eventually enrolled in college the following year(s) was not included. A ccordingly to the literature (Horn et al 2006), many Latina/o students delay college enrollment up in till their late twenties as opposed to enrolling in college immediately after high school The most recent follow up of ELS:02 may have this information
95 available unfortunately, du e to tim e constraint I did n ot have sufficient time to access and analyze the most current follow up of ELS:02 Table 3 1 8 Frequency of c ollege e nrollment (n=2209) College Enrollment Enrolled in College Frequency Percent Yes 1179 53% No 668 30% Independent Variables model allowed me to organize the independent variables based o n the layers of the model. Table 3 19 summarizes the independent variables used in this study. Table 3 19 Summary of independent variable s and indices Items ELS:02 Source Variable Variable type Scale Gender BYS14 Dichotomous dummy 0 = Male (refer ence), 1 = Female level of education BYPARED Ordinal Categorical dummy Recoded: 1 = High sc hool diploma or less (reference ), 2 = Some college but no degree, 3 = 4 = degree, 5 = Advanced degree. Family comp osition BYFCOMP Categorical dummy Recoded: 1 = Both parents (reference), 2 = one parent & one guardian or two guardians, 3 = Mother only, 4 = Father only, 5= One guardian only, 6 = Parent or Guardian lives with student less than half the time. Num ber of in home siblings BYSIBHOM Categorical dummy Recoded: 1 = Zero (reference), 2 = One two, 3 = Three four, 4 = Five or more language BYHOMLN Categorical dummy Recoded: 0 = English (reference), 1 = Spanish, 2 = Other fluency BYPLANG Categorical dummy 1 = Not fluent (reference), 2 = Partially fluent, 3 = Fluent, 4 = Fluency unknown, 5 = Native English speaker Generational Status BYGNSTAT Categorical dummy Recoded: 0 = Not American born (reference), 1 = First Gener ation, 2 = Second Generation.
96 Table continued. Items ELS:02 Source Variable Variable type Scale Family income BYINCOME Ordinal Categorical dummy Recoded: 1 = $35,000 or less (Low) (reference), 2 = $35,001 $50,000 (Middle), 3 = $50,001 $75,000 (Mid high), 4 = $75,001 or greater (High) Do you plan to continue your education right after high school or at some time in the future? BYS57 Categorical dummy Recoded: 1 = Yes, after high school (reference); 2 = Yes, after one year; 3 = Yes, over one year; 4 = Yes, but plan to continue my education. What type of college attend BYS58 Categorical dummy Recoded: 1= Four year college (reference), 2 = Two year college, 3 = Vocational School Social network s (Mother, Father, Friend, Close relatives, Guidance Counselor, Favorite Teacher Coach) desire for 10 th student to do after high school BYS66 Categorical dummy Recoded: 1 = Attend college (reference), 2 = Get a job, 3 = Attend trade school, 4 = Join the military, 5 = Get marry, 6 = Whatever the student wants to do Social networks (Mother, Father, Friend, Close relatives, Guidance Counselor, Favorite Teacher Coach) desire for 12 th student to do after high school F1S44 Categorical dummy Recoded: 1 = Attend college (reference), 2 = Get a job, 3= Attend trade school, 4 = Join the military, 5 = Get marry, 6 = Whatever the student wants to do Number of colleges applied to F2BO3_P Categorical D ummy 1 = One (reference), 2= two, 3 =three, 4 = four, 5 = five, 6 = six, 7 = seven, 8= eight, 9 = nine, 10 = ten, and 11= eleven or more. High school urbanization BYURBAN Categorical dummy 0 = Urban (reference), 1 = Suburb, 2 = Rural High school type BYCTRL Categorical dummy 0 =Public (reference), 1 = Catholic, 2 = Private college choice is one of the theoretical frameworks guiding this study. Figure 2 1 represents how the in dependent variables used in this study come s the rationale for choosing the selected independent and dependent variables for t his study. To
97 conduct the necessary analyses for this study, many of the independent variables were recoded into dummy variables The reference group for each variable is in parentheses Layer One: The Individual Habitus The most prominent layer addressed in this study is layer one: individ ual habitus. social networks, layer one address es a variety of other key demographic and social and cultural characteristics. The variables addressed in layer one ar e often used in research stages. The following sections describe more specifically those independent variables Demographic Characteristics Previous research studies have examined the impact of gender and women are more likely to enroll in college tha n Latino males. However, s ome studies have found that Latino males are more likely to go off to colleg e, whereas Latina women attend local college s (Zarate & Gallimore, 2005). The gender gap continues to be evident for every racial group but i t i s especially prevalent among low inc ome and studen ts of color (King, 2006). According to Saenz & Ponjuan (2009), Latino males have the lowest degree attainment among all minority males, including African American males. Despite these disparities in college enrollment, differences in the coll ege choice process of Latina/o students based solely on gender are not always differences in the college choice process of Latina/o males (reference group) and females
98 Race college choice process. Research studies suggest that studen ts of different racial/ethnic backgrounds experience the college choice process differently from one another. For examp le, some researchers suggests that White and Asian students are more likely to successfully navigate the college choice process and enroll in more prestigious institutions (Snyder & Dillow, 2009) in comparison to Black and Latina/o students. In particular, a smaller percentage of Black a nd Hispanic students enroll in postsecondary institutions compared to their White peers ( L. Perna, 2006). However, I am not interested in comparing the college choice process of different racial groups among each other Ther efore college choice process and enrollment rates of the four dominant racial/ethnic groups in the United States: White, Black, Latina/o, and Asian. O nly students who identified as Hispanics w er e examined for this study. Other variables that were not ere provided in composition and the number of in home siblings For example, hous eholds that have two working parents are more likely to earn more money than households with on ly on e working parent. A L atina/o hou sehold of four that has two working parents is more likely to tuition fee s compared to a single pa rent household of f our family members As previously discuss ed siblings can also be very influential in the college choice process of Latina/o students (Hurtado & Carter, 1997; Gandara, 1995). number of in home siblings analyzed to see whet her siblings were significant in college choice process
99 Social Capital According to various studies (Ceja, 2006; Cohen, 2009) the amount of social capital students obtain can greatly impact their college choice process and their decision to attend college. In the educational context, social capital refers to the ability social networks can facilitate educational advancement for students (Bourdieu, 1986; Stanto Salazar, 2004) whether through educational agents, friends, and/or family members. First generation students often lack the social capital that many of their non first generation peers experience since their parents lack first hand experience with t he college choice process (Inkelas & McCarron, 2006). There fore, fi rst generation students greatly benefit from the relationships they establish with the ir institutional gatekeepers because many are unfamiliar with the requirements of higher education and need additional assistance in navigating the c ollege choice process Institut ional gatekeepers provided more than resources to students, they provided valuable information, norms, and support (Coleman, 1988; Bryan et al 2011). W hen it comes to obtaining college source of social capital for low income students and students of color who are first (Bryan et al., 2011). For this reason, this study is looking at various social networks that may process. The primary variable connected to social capital asked Latina/o st udents what the following individual s mother, father, friend, close relative, school counselor, favorite teacher, and coach think is the most important thing for them to do afte r h igh school. The variables, BYS66 and F1S44, was not recoded. The only difference between BYS66 and F1S44 these two variables is
100 timing. T he first question was asked in 2002 (BYS66 ) and the second variable was asked in 2004 (F1S44) Cultural Capital Cultural capital refers to general characteristics, such as skills, language, cultural knowledge, and manners that t ypically social class status ( L. Perna, 2006; Bourdieu, 1986). Many researchers have suggested that accessed to the type of social and cultural capital needed to effectively navigate the college choice process ( L. Perna & Titus, 2005; L. Perna, 2000). Key variables that s generational status (BYGN STAT PLANG ), and native language (BYHOMLN ). College educated parents, along with having information in their homes about college ( L. Perna, 2006), is one form of social capital that may help to promote college enrollment. Studies have proven that first of ten encounter additional challenges when navigating the college choice process and graduating from college compared to students who have college educated parents (Ceja, 2001; Choy 2001). Therefore, the ELS:02 variable BYPARED was analyzed to compare the colleg e choice process and enrollment rate between first generation Latina/os and Latina/o s who have college educated parents. Latina/o students generational status, in particular whether students and their parents are U.S. born citizens or not may impact the ir educational experiences and opportunities. For example, t he state of Florida is currently debating whether U.S. born students whose parents are undocumented should continue to pay international tuition fee s (Hickson, 2012). Therefore, Latina/o students whose parents are either U.S.
101 citizens or permanent residents, have an advantage in terms of receiving federal funding and paying in state tuition vers u s students whose parents are undocumented. As for undocumented students, only 17 states have provisions allowing for in state tuition rates for these students California, Colorado, Connecticut, Illinois, Kansas, Maryland, Minnesota, Nebraska, New Jersey, New Mexico, New York, Oregon, Texas, Utah, Washington, Oklahoma, and Rhode Island either through Boar d of Regents decision s or state legislation (Morse, 2013). Two of the three states with the largest Latina/o population, California and Texas, currently allow undocumented students to receive state financial aid (Morse, 2013) w hereas A rizona, Georgia and Indiana sp ecifically prohibit in state tuition rates for undocumented students, and two states Alabama and South Carolina prohibit undocumented students from enrolling at any public postsecondary institution (Morse, 2013). Additionally, Latina/o stude nts whose parents were either born or raised in the United States (i.e., third generation) may have an advantage over first or second generation Latina/o students when it comes to navigating the K 12 educational system since their parents are familiari zed with the American s chool system. While f irst generation Hispanic immigrant youth have made tremendous strides in educational attainment relativ e to their parents, this trend of upward mobility reverses by the third generation (Perreira, Harris, & Lee, 2007 ). Contrary to popular belief, second generation Latina/os (i.e., U.S. born with at least one immigrant parent) are more likely to aspire and attend college in comparison to third generation Latina/o students (Perreira, Harris, & Lee, 2007). Nonetheless, s econd generation Latina/os are not graduating from college at the same rate as their native non Hispanic White peers (Fry, 2007).
102 Moreover s another factor that fa l l s under cultural capital. Latina/o parents who are native En glish speakers or fluent in English are better able to communicate efficiently with teachers, administrators, and college Some i mm igrant parents are unable to help their children with school work due to their limited English skills, knowledge of the U.S. educational system, and access to bilingual institutional support f rom their (Cruz, Capps, Vericker, & Kuehn, 2009). Consequently, many immigrant parents of low socioeconomic status often rel school administrators and teachers to make decisions about what is best for their child education wise (Cruz, Capps, Vericker, & Kuehn, 2009). However, Cruz et al (2009) f ound that the absence of English spoken in Hispani c households is more detrimental to first generation immigrants in comparison to their second generation Latina/o counterparts Second generation Latina/os have higher educational attainment than third generation Latina/os even whe n they live in non Englis h households (Cruz, Capps, Vericker, & Kuehn, 2009). especially, as it relates to For example, recent immigrants may face more challenges in scho ol compared to native English speakers in terms of understanding the curriculum, taking st andardized exams or being admitted into advanced placement or honor programs due to their limited fluency in English. This is especially true for students who migra ted to the U.S. at an older age. Thus, native English speakers or fluent speakers are more likely to have an advantage over students who are partially or not fluent in English when it comes to navigat ing the college choice
103 process. Therefor e, the ELS:02 va riable labeled native language was selected because it asked students what was their first spoken language The variable was recoded to include only three options ; English (reference), Spanish and other. Supply of Resources Another category selected fro supply of resources This c onsist of family income and any unmet need the student may have had after all grants ha d been disperse. However, only family income was pro vided in ELS:02 to measure for supply of resources Research suggests that students from low income levels navigate the college choice process differently from students who come from wealthier family backgrounds (Gloria & Segura Herrera, 2004). St udents from more affluent family backgrounds are more likely to obtain the nece ssary information and resources needed t o successfully navigate the college choice process and enroll in college. Accordingly, choice process and enrollment rate of Latina/o students among four groups: low income, middle class, middle high income, and high income class. Research also shows a positive relationship between family income and the number of college applications submitted (Hurtado et al 1997). The variable F2BO3_P was included to investigate how the college choice process of Latina/o students differ between students who applied to only one college versus students who applied to more than thr ee Expected Costs refers to both the cost of attending a postsecondary institution and the lost wages of attending college For this study, only cost of attendance is being measured by using ELS:02 v ariables type of institution students plan to attend and college. Cost of
104 attendance, especially for financially disa dvantage students is found to be o ne of the main reasons students prefer red to attend community colleges over four year institutions due to l ower tuition rates S. Hurtado et al (1997) found th at was highly associ ated to the type of institution students considered to attend. Therefore, ELS:02 v ariable BYS58 w as analyzed to investigate whether the college choice process of Latina/o students differ between students who aspired to a ttend a four year college (reference) versus a community college or vocational school. A long with their likelihood of actually enrolling in college However, the pathway to degree attainment for Latina/o st udents beings with college enrollment T herefore, the variable BYS57 was examined in o rder to determine when students plan ned to enroll in college. Another important factor in layer one conceptual model is d emand for higher education (i.e., academic preparation and a chievement). However, th is is a broad construct previously addressed but it is diff icult to measure academic preparation and achiev ement since each state and school district ha s their own grading scale. Nonetheless, it is still important to review how demand for higher education impacts the college choice process and decision to enroll in college of Latina/o Layer Two: School and Community Context ; the school and community context. The following s ections describe more specifically those independent variables that were model. School Location Urbanization As supported by the literature, the type of high school Latina/o students attend are critical to their college choice process. For example, urban schools have unique
105 physical and demographic characteristics that differentiate from suburban and rural school districts. Unlike suburban and rural school districts, urban school s operate in densely populated area s serving significantly more students (Kincheloe, 2004). As a result, urban school districts are frequently marked by higher concentrations of poverty, greater racial and ethnic diversity, larger concentrations of immigrant populations linguistic diversit y, and more frequent rates of student mobility (Kincheloe, 2004, 2010). This impacts the overall q uality of education s tudents receive in urban s ettings compared to students who attend schools in the suburban. Other researchers, have found that u rban schoo ls often fail to provide environments of high academic expectations for their students (Griffith, 2000; Noguera, 2003; Vale n cia, 2000; Valenzuela, 1999) Nonetheless, r u ral scho ols also have their limitations. For example, several student characteristics including large number of students with special needs, limited English skills, and lower college enrollment rate s from this student population, often add to the difficulty of recruiting and retaining high quality and experienced teachers (Monk, 2007). Simi lar to many urban school districts, rural districts are seriously impoverished. Conversely, rural schools are distanced from highly populated areas which makes it di fficult to hire teachers who live in the city s chools have very high teacher turnover, a lack of teacher diversity, and a l arge concentration of children of migrant w orkers Many of the teachers at these schools lack proper training to p roperly teach such population s As proven by s tudents who attend schools with high concentrations of low income Black and Latino students are more likely to have inexperienced or unqualified teachers, fewer
106 demanding college preparatory courses, more remedial courses, and high er teacher turnover. Th er efore, the ELS:02 analyze Latina/o ased on the type of school s tudents were enrolled in; urban suburban or rural school. I also considered including the pe rcentage of students that receive free or reduced lunch (BY10FLP) as an indicator of socioeconomic status of the school ho wever, recent research suggest High School Type Further more Latina/os have been underrepresented in academically rigorous enrichment programs and courses, such as Gifted, Magnet, Honors, Advanced Placement [AP] and International Baccalaureate [IB] (Gandara & Contreras, 2009), reducing their probabilities of compet ing in the college pool of selective institutions According to Brown, Kurzweil, and Tobin, (2005) the more selective institution are the higher the probability students will be retained as a result to the various resources and programs t hey offer to their students in co mparison to community colleges. Other institutions to their disproportionate enrollment rates at disadvantaged, overcrowded, and low achieving K 12 schools (Fry, 2 005; Aguirre & Martinez, 1993). Many of these schools have high student to teacher ratio s and fewer resources to promote and encourage college enrollment (Perna & Titus, 2005). Therefore, the second variable to h de termines if a high school i s private, public, or Catholic. The type of high s chool students attend i s important to analyze since it often influences the ir college choice process and their decision to enroll in college.
107 Layer Three: Higher Education Contex t Another relevant layer addressed in this study is layer th ree: higher education context. This layer addresses how certain institutional characteristics impact the college choice process of Latina/o students. Before applying or enrolling in a particular postsecondary institution, many students take into account certain institutional characteristics such as: (a) program of study (b) reputation of school (c) cost (d) location (e) personal or family reasons (f) or another reason. Students were also asked in the 10 th grade, what type of college (i.e., two year, four year, or vocational) they planned to attend after high school. This is important to know because by the 9 th grade many students are already thinking about college (A.F. Cabrea r and La Nasa, 2000b) Institutional Characteristics According to the literature and findings, some of the most influential postsecondary characteristics to Latina/o students when deciding which college to attend are: location, program of study, and financial reasons (Kurlae nder, 2006; Fry, 2006; Burdman, 2005). Latina/o students main reason to attend a community college has to do with their c lose proximity to home and lower tuition cost in comparison to four year institutions (Fry, 2006). Since many Latina/o students have c lose ties with their families many decide d to live at home while attending college This allows students to fulfill familial obligations of caring for family members or contributing to bills Since community colleges are more accessible to Latina/o student s compared to st ate universities many prefer to enroll at local two year colleges (Kurlaender, 2006). According to Rooney (2008), students are more likely to enroll in the college s that offer them the most financial aid. Another important institutional cha racteristic Latina/o
108 students take into account is the program of study. This is critical for students who know what they want t o major because not every institution offers every major. Methodological Considerations Furthermore, it is important to recogniz e that there are certain methodological considerations which must be addressed that are relevant to this particular study. Understanding these challenges will help to limit any data analys i s errors and provided the strongest possible results. A primary met hodological concern was the issue of using a secondary dataset. The information needed to answer the research questions was gathered despite the fact that ELS:02 survey questions did not exactly match this fter all ELS:02 data was not collected specifically for this study. Existing research highlights the importance of addressing each of these challenges ( Cellini, 2008; Dowd, 2008; Hahs Vaughn, 2007; Thomas & Heck, 2001 ). The following subsections go into further detail about the two methodological considerations that need to be taken into account for this study. Analyzing Secondary Datasets For this particular study, the primary methodological concern is the issue of using a secondary dataset. Within the past decades, second ary datasets have become more popular in higher education research (Thomas & Heck, 2001) due to the large number available and the opportunity to use a robust dataset inexpensively (Strayhorn, 2009). When using NCES restricted data files, such as ELS:02, i t is imperative for researchers to understand the design effects and weights used in this dataset to insure it is properly analyzed (Hahs Vaughn, 2007). Despite the benefits of using secondary datasets, there are a variety of disadvantages in using such da tasets (Perna & Titus, 2005; Thomas & Heck, 2001). One disadvantage is the oversampling of particular characteristics in
109 national datasets (Hahs Vaughn, 2007; Hahs Vaughn, et al., 2011; Strayhorn, 2009) and another is homogeneity. Lacking homogeneity may resul t in incorrect standard errors causing Type I errors. To address the non independence issue causing hom ogeneity, it is helpful to consider using a normalized weight and specialized software (Hahs Vaughn, et al., 2011; Thomas & Heck, 2001). I opted to use a normalized weight from the weights provided by the ELS:02 dataset since the sample only includes Latina/o students and not the entire student population. Accordingly, R Studio and M Plus software were utilized for the purpose of this study. R Studio software is designed to produce correct estimates for data collected through complex sampling me thods and M Plus was used for its ability to conduct multivariate mult were chosen over other existing programs (e.g. SAS, Stat a, Sudaan, SPSS, S Plus) due to the Studio and M Plus. Most of the NCES datasets do not use simple random sampling techniques ; therefore weights were used to ensure the data are representative of the intended population (Hahs Vaughn, 2007). Weights are applied during data analysis to minimize the disproportionate influence of the cases or subjects that were oversampled in the complex sampling design (Thomas & Heck, 2001). Failure to apply the weights completely disregards the complex survey sign and will result in analyses that reflect s the sample and not the larger population of interest. Analyzing NCES data without using the weights can also violate the independence assumption and generate parameter estimates that are biased (Thomas & Heck, 2001). The weights provided in ELS:02
110 restricted data file s are intended to compensate for the unequal probability of selection of secondary institutions and students that comprise of the final sample in this data In addit ion, these weights adjust for multiplicity at the institutional and student levels, unknown student eligibility, nonresponse, and post stratification (Wei et al., 2009). The EL S:02 methodology report provides detailed information abo ut the specific purpose and use of each weight included within the dataset. The methodology report and existing research studies (see Hahs Vaughn, 2005 & 2007; Thomas & Heck, 2001) were utilized to ensure the normalized weight for the sample were calculated correctly and applied accurately during data analysis. As mentioned previously in C hapter 3 this normalized weight will be used to conduct all of the statis tical analysis for this study. Estimating Causal Effect with Secondary Data A large percent of higher education research ers rely heavily on secondary data and should be aware of the statistical challenges and limitations inherent in using these types of data to claim causality (Titus, 2007). A primary statistical problem in higher education research is self selection bias ( Dowd, 2008; Titus, 2007). Self selection consists of one or more independent variables included in the statistical model as 2001). If this correlation is not prope rly controlled during the stat istical analyse s it can result in model misspecification and inaccurate coefficient estimates (Titus, 2007). By using large secondary datasets, such as NCES and NSF, omitting variable bias is a statistical concern that can of ten result in self selection bias. When an independent variable that might influence the outcome variable is not included in the regression model, omitted variable bias takes place (Cellini, 2008). A benefit of using secondary national datasets is the weal th of variables it provides. However, t hese
111 datasets may not provide enough information about complex constructs to allow researchers to disentangle the effects of student characteristics that may led them to make certain choices about college (Dowd, 2008; Perna & Titus, 2005). If these variables are not taken into consideration in the logistic regression models, omitted variable bias may result in unreliable estimates because one or more independent variables in the model are correlated with the error term In this study, the ELS:02 dataset prov ides countless variables to properly demographic and institutional variables used in this study have been examined by other college choice studies The variables representing the college choice stages of Hossler college degree aspiration s ( predisposition ); seeking college information via way of social networks (search); sele cting an institution based off certain reasons (choosing a college); and a ttending college (enrollment). Analytical Methods Furthermore, this section provides an overview of the analytic methods used to address the research questions guiding this quantitative study. The data analysis consisted of two st ages, preliminary and advanced. The initial descriptive findings of these a nalyses were presented in C hapter 3 in conjunction with the sample presentation The more advanced analyses will be presented in C hapter 4 Quantitative methods w ere used to address the research questions in th is study. The data analysi s includes f ive logistic reg ression models which allow ed for all the research questions to be addre ssed. As a result to having multinomial and dichotomous categorical dependent variables and multiple independent variables, the most appropriate statistical
112 t echnique to use in this study i s logistic regression. It is recommended to use logistic regressio n in studies where the outcome variable is categorical and the independent variables are categorical or continuous (Peng, Lee, & Ingersoll, 2002). The preliminary analysis in C hapter 3 highlight key descriptive statistics and frequencies between independe nt variables. Descriptive statistics were utilized to descriptive analyse s was used to examine the frequencies of the dependent measures in this study. These descriptive statistics helped identify the percent age of students who aspire to obtain a college degree, who they were more likely to seek for college information, the primary reason they selected an institut ion, and whether they enrolled in college. Certain variables were reco rded in order to comply with this Afterwards, multiple imputations w as performed since there was more than 5% of missing data in the variables. To perform multiple imputations, we used the function mice of the package mice (Buuren & Groothuis Oudshoorn, 2011) on the software R version 2.15.1 (R Core Team, 2012). The ad vanced data analysis includes five multivariate multinomial models, which allo wed for all the research questions to be addressed. The first research question investigates whether there is a correlation between the ex have for Latina/o students and the ir aspiration to attend college, which was a multilevel categorical variable. The use of the second and third dependent variable s allow ed for this question to be sufficiently answered in terms of fully understanding which social networks are most influential in the college choice process of Latina/o students.
113 The second research question used a dichotomous categorical variable This question sought to identify wh ich sources Latina/o students we re most likely to seek for college information in both the 10 th and 12 th grade based on the expectations their social networks ha d for them. The third research question used a dichotomous multinomial categorical variable since it sought to identify whether there is a correlation their main reason for choosin g a college. The final research question goes beyond the college choice proc ess rate s The fourth research question used a dichotomous categorical variable that identifies if there is a correlation between the expectat their decision to enroll in college after high school. Hierarchical generalized linear models will be applied to both the binary and categorical variables using M plus 7.0 (Raudenbush & Bryk, 2002). M pl us was primary used for its ability to analyze multilevel data and take into account the complex structure of the data especially, since this data has weights. Limitation of Study Like any other study, there are several limitations that deserve further a ttention. For one, ELS:02 only follows students from 10 th grade and two years after their high school graduation at least for this version However, it excludes 9 th grade which according to some researchers consist of a crucial transition year, especially towards r a and La Nasa, 2000b). T his study can only evaluate students two y ears after high school thus, r esearchers interested in study ing students college graduation rates will have to wait for ELS:02 third follow up release whic h was conducted in 2012 but w as not a vailable in time for the completion of this study.
114 Another limitation of this study relates to the self reported nature of many of the variables used during the data anal ysis portion. These self reported variables come from the students and may c ause bias in statistical analyse s because as human beings individuals may not always re call accurate estimates or may answer based on what they think is expected of them. Therefore interpreting the variables that are self reported must be done under a nuanced perspective and acknowledge t hat these items are sel f fore, using secondary data limits the application of the se variables to the theoretical frameworks used for this study The selected survey questions were carefully chosen and modified to appropriately answer the research questions for this study. Although the ELS:02 dataset was the most robust and current dataset available at the time of this study, it was not a perfect match since the survey questions were not created to fit the research questions in this study. Contributions of Study This study contribut es valuable information to the existing research literature on ollege choice process based on their relationship with specific social networks. Adaption of these two models for the purpose of this study allow ed for the examination of key variables (e.g., demographic characteristics social/cultural capital, high sch ool context, institutional characteristics) that typically have been absent from existing studies o n The identification of these variables can result in a better understanding of the factors associated with Latina
115 Second, this study utilized the most recent nationally representative data to i dentify the characteristics of Latina/o students who experience the college choice process and actually enrolle d in college. Equipped with a better understanding of the co llege choice process of Latina/o students, policymakers, school administrators, and higher education leaders can begin to develop strategies and programs aimed at iliarity with college requirements, financial aid literacy, and college enrollment rates at four year institutions. After all, effective public policies and institutional practices will only help increase the number of first generation Latina/o students at four year institutions. In addition to improving the economical dilemma of this country, reducing government dependency and reducing unemployment rates, higher education will also improve the lifestyle and well being of Latina/o s Table 3 20. Summary of multivariate logistic regression models Independent Variables Model 1 Predisposition Stage Model 2 Search Stage Model 3 Search Stage Model 4 Choice Stage Model 5 College Enrollment Demographic Characteristics Gender X X X X X Family composition X X X X X Number of in home siblings X X X X X Social & Cultural Capital X X X X X X X X X X X X X X X X X X X X Social Networks expectation for 1 0 th grader after high school X X X X X Social Networks expectation for 12 th grader after high school X N/A X X X Supply of Resources Family income X X X X X Number of schools applied to N/A N/A N/A X X Expected Costs Type of College X X X X X When students plan to enroll in college X X X X X High school Context High school type X X X X X Degree of urbanization X X X X X
116 CHAPTER 4 DATA ANALYSIS AND RESULTS The objective of Chapter 4 is to present the data a na lysi s results that were conducted to address the four research questions guiding this study. Results from the descriptive analysis were presented in Chapter 3. The advanced data analysis section presents results from the logistic and multilevel regression models. It concludes with key findings from the data analys i s. Advanced Data Analysis This section will present results from the advanced data analysis used to address the four research questions guiding this study. First, an ordinal multilevel logisti c (2006) modified conceptual model. Next, two multilevel logistic regression models were us ed to examine who Latina/o students were most likely to go to for college entrance information as 10 th and 12 th graders. Afterwards, a multinomial model was used to examine the main reason Latina/o students chose their college. Lastly, a binary logistic mo results of each regression model are discussed in the following subsections. A multilevel ordinal l ogistic regression model was used to address the first research question guiding this study. For this question the dependent variable of college aspirations and any of the i conceptual model. Overall, the model was significant and it predicted 66.9% of Latina/o
117 The first block of variables examine the probability that students would aspire for a c ollege degree as a function of the following demographic variables: gende r, family composition, and the number of in home siblings. However, not found to be s However, moving from male to female students, the log odds of Latina/o students aspiring for a college degree decreased by 0.227. Thus, Latino males were more likely than Latinas to a spire for a college degree. The second block of variables examined the probability that students would aspire for college degree based on social and cultural capital (Table 4 1). Social capital was measured by using the expectations social networks had fo r students. T he results 1). Cultural capital was measured one unit change in Latina/o pare aspiring for a college degree increased by 0.168. Hence, Latina/o students whose
118 parents were native English speakers were more likely to aspire for a college degree than students whose parents did not speak English. The third block of variables pertains to supply of resources as measured by family income and the number of institutions students applied to. The type of college 10 th graders planned to attend and when they intended to enroll was used to measured students expected costs (Table 4 1). Despite the findings of other research studies, predisposition stage. On the other hand, the number of college applicati ons 12 th grade aspirations. In particular, for every additional school Latina/o students applied to, the log higher decreased by aspirations for a college degree. Perhaps those students who applied to only one or two schools were more confident and convinced about the school t hey wanted to get in versus students who applied to more than three colleges. Additionally, the two variables used to measured students expected costs were aspirations (Table 4 1). In particular, when students planned to enroll in college was when students planned to enroll in college, the log odds of Latina/o students aspiring for a colle ge degree increased by 0.175. Ironically, Latina/o students who planned to postpone school after a year or more were more likely to have higher degree aspirations than Latina/o students who planned to enroll in college after high school. Secondly, the
119 type predictor of students college aspirations. For every one unit change in the type of college Latina/o students planned to enroll in, the log odds of Latina/o students aspiring for a college degree decreased by 1.364. As expected, Latina/o students who planned to attend a four year college had higher degree aspirations than Latina/o students who aspired to attend a two year college or vocational school. The last block of variables entered into the regression model examined Latina/o urbanization of the school and the type of school (public, private, or Catholic) students attended (Table 4 1). According to th found between school type a nd students degree aspirations. Multilevel Logistic Regression: Search Stage in 10 th grade Linear multivariate regression was used to address the second research question guiding this study. The dependent variable of interest consist of the most frequent sources (i.e., parents, guidance counselor, friends, college representatives etc.) 10 th grade Latina/o students seek for college entrance information. To reduce reputation, only those independent variables found significant for each source are included, in stead (2006) conceptual model. The regression model used in this study explained that 45% of the Latina/o student population went to their guidance counselor for college e ntrance information.
120 (Table 4 2). For every one unit change in family composi tion, the log odds of Latina/o students going to their guidance counselor decreased by 0.024. Thus, Latina/o students who came from a one parent household versus a two parent household, were less likely to seek college information from their guidance couns elor. For every additional school Latina/o students applied to, the log odds of Latina/o students seeking college information from their guidance counselor increased by 0.011 (Table 4 2). Thus, the more schools Latina/o students applied to the greater the likelihood of them seeking college information from their guidance counselor. The regression model reported that fewer (38%) Latina/o students went to their teacher for college entrance information. There was only one significant variable found among all the independent variables in the model (Table 4 10 th th student, the log odds of s tudents seeking college information from their teacher increased by 0.149. Thus, students whose fathers desired for them to do as they please v erses going to college were more likely to seek college information from their teachers. Fewer (15%) students went to their coaches for college information. However, this variable was found much more significant in relationship to other independent variab college the y planned to attend (Table 4 difference in the probability of Latina/o males and females seeking college information from their coach. Moving from male to female, the log odds of Latina/o students seekin g college information from their coach decreased by 0.582. Latino males were more likely
121 to seek college information from their coach than Latinas. For every one unit change in college information from their coach decreased by 0.193. Thus, Latina/o students whose coach desired for them to attend college were more likely to seek college information from them. Further on, the type of college Latina/o students planned to enroll in w as information. For every one unit change in the type of college, the log odds of Latina/o students who seek college information from their coach decreased by 0.955. More sp ecifically, Latina/o students who planned to attend a four year college were more likely to seek college entrance information from their coaches than students who planned to attend a vocational school. As expected, many more Latina/o students (47%) went t o their parents for info rmation from their parents ( Table 4 3). In particular, La tinas were more likely to seek college information from their parents. Ironically, t he higher the level of education obtained by the parent, the less likely students were to seek information from their status. For every one unit change in generational status, the log odds of Latina/o students seeking college information from their parents decreased by 0.108 ( Table 4 3).
122 Immigrant students were more likely to seek college information from their parents than second generation Latina/os. Approximately, 4 3% of 10 th graders went to their friends for college entrance in the probability of Latina/o male and female students seeking college information from their friends. Moving from male to female, the log odds of students seeking college information from their friends increased by 0.357 ( Table 4 attend a four year, two year, or vocational school (type of college) was also significant students who planned to enroll in a vocational school were less likely to seek college information from their friends than Latina/o students who planned to enroll in a four year institution. On the other hand, fewer students (28%) reported to obtained co llege information from their siblings. Only two statistically significant variables were found. This includes the type of college students planned to attend and their Table 4 3). For every one unit change in college type, the lo g odds of Latina/o students seeking college information from their siblings decreased by .0529. Students who planned to attend a four year college verses a vocational school were more likely to seek out their siblings for college information. Another signi had students and their decision to go to their siblings for college admissions information. Contrary to their siblings, Latina/o students were more likely (30%) to go to their relati ves for college information. A statistically significant difference was found among
123 probability of students acquiring college entrance information from relatives. Moving from male to female, the log odds of Latina/o students seeking information from relatives increased by 0.375 (T able 4 4). Latina/o students who planned to attend a two ye ar college or vocational institution and planned to attend college after high school were less likely to seek college information from their relatives (Table 4 4).For every information from their relatives increased by 0.084. Furthermore, the probability of 10 th grade Latina/o students seeking college information from college sources were as follow: (34%) gather information from college publications, (24%) from college repr esentatives, and (33%) from college search and female students decision in obtaining college information from college publications ( Table 4 4), along with their parent the level of education of the parent the less likely students were to seek college information from college publications. The number of college applications students applied to were statistically signif icant in students decision to obtain college entrance ( T able 4 type of college students om college representatives ( Table 4 4). As for students who did not go to any of these sources (24%), only four independent variables were found to b e statistically significant ( Table 4 5). Latino males
124 were more likely not to seek college information from any of the other sources compared to Latinas. The number of college applications Latina/o students applied for was statist from any of the above sources. When students planned to enroll in college and the type to their decision not to seek college information from any of these sources. For every one unit change in the type of college, the log odds of students not seeking college information from any source decreased by 0.875. However, for every one unit change in when students planned to enroll in college the log odds of them not seeking information from any of the others sources increased by 0.267. Multilevel Logistic Regression: Search Stage in 12 th grade rch stage as 10 th and 12 th graders. However, for this model we are looking at those sources students were more likely to seek for college information as 12 th graders. The previous model looked at the sources students seek as 10 th graders. The sources consi st of school counselors, teachers, parents, family, friends, libraries, college sources, and none of the above. Only those independent variables found significant for each source are included in this section. The regression model used in this study explai ned that 62% of 12 th grade Latina/o students went to their guidance counselor for college entrance information. This is a much higher rate compared to their 10 th grade results. According to the findings, there were many significant variables related to stu
125 from their school counselor ( Table 4 status, the log odds of students going to their sc hool counselor decreased by 0.326 (Table 4 6). In particular, the number of college applications students applied to as 12 th information from their guidance counselor. Howev er, for every one unit change in school desire for the student, the log odds o f students going to their 12 th grade school counselor decreased by 0.206. The regression model also reported that fewer (46%) Latina/o students went to their teacher for college entrance information. Only one independent var iable was found significant ( Tab le 4 6); the number of college college applications, the log odds of 12 th grade students seeking college information ore college applications students applied to the more likely they were to seek college information from their teacher. Moreover, fewer Latina/o students (39%) went to their parents for college entrance information as 12 th graders than as 10 th graders. Non etheless, there were more s ignificant variables found ( Table 4 desire for 10 th th More specifically, non nati ve English speakers were less likely to go to their parents for
126 college entrance information compared to native English speakers ( Table 4 6). For college information from the ir parents increased by 0.256 ( Table 4 6). In term of students seeking information from their p Table 4 6). As for y one unit change the log odds of students obtaining college college information fr 38% of Latina/o students reported to having gone to their family members for college information. The only significant to obtain college information from family members as 12 th graders (Table 4 7). More specifically, for every one unit increased in generational status, the log odds of students going to their family for college informatio n decreased by 0.291. A similar rate (45%) of 12 th grade Latina/o students went to their friends for college entrance information compared to when they were 10 th graders (43%). There of 12 th grade Latina/o male or female students seeking college inf ormation from their friends ( Table 4 7). Comparable findings found that 12 th grade Latinas were more likely than Latino males to go to their friends for college information. The second sign
127 th decision to go to their friends for college information. For every one unit change in ts seeking college information from their friends increased by 0.159. The probability of 12 th grade Latina/o students seeking college information from any college source such as, college publications, college representatives, and college search guides wer e the second highest (61%) source, next to guidance counselor. These three variables were recoded into one variable called college sources. A taining college info rmation from college sources ( Table 4 7). In particular, Latinas were more likely to obtain college information from college sources .001), and teachers th grade students was found significant in their decision to seek college knowledge from different college sources (Table 4 7). More students, the log odds of them obtaining postsecondary information from college sources increased by 0.238 and 0.090, respectively. On the other hand, for every one th grade students, the log odds of them obtaining information from college sources decreased by 0.236. The number of college Lastly, the type of college students planned to attend was found statistically significant in their decision precisely, students who planned to attend a vocational school were less likely to pursue
128 college information from college sources than students who planned to attend a four year college. A new variable that was included in this model but was not made available school libraries, and college libraries as a source students used to pursuing college inform ation. Approximately 20% of Latina/o students went to libraries for college conceptual model was found to be statistically significant in testing stu search stage ( T able 4 8). As for students who did not go to any (21%) of these sources for college entrance information, only three independent variables were found to be statistically significant ( Table 4 8). The desire 12 th ers college information from any of the above sources. In particular, for every one unit ormation the log odds of students not seeking college information from any of the above sources increased by 0.372. The number of college applications Latina/o students a pplied for applications, the log odds of Latina/o students not seeking college information from any of the previous sources decreased by 0.160, which is more than what i t decreased when students were in 10 th grade (0.087). Multilevel Multinomial Logistic Regression: Choice Stage Multinomial logistic regression was used to test the choice stage, as it relates to the main reason they selected their chosen college. The ma jority of Latina/o students
129 chose college location (28%) as their main reason of attending their chosen college, followed by other reasons (22%). The four remaining reasons students selected their chosen college were: program of study (19%), affordability (10%), and family reasons (10%). There were a few variables found significant in this model. This includes the type of college students planned to attend and the expectations social networks had the as 10 th and 12 th graders. Mor e specifically, for every one unit increase in the type of college Latina/o students planned to enroll in there was a 0.302 increased in the relative log odds of students selecting their chosen college based on reputation verses college programs (Table 4 9 ). There was no significant relationship found between any independent variables and whether students selected 9). nd between the type of college students planned to enroll in and their decision in selecting their chosen college based on location versus college programs (Table 4 10). Furthermore, a significant relationship existed between the desire Latina/o th th graders (Table 4 10). A one 10 th grade Latina/os resulted in a 0.170 increased in the relative log odds of students s electing their chosen college based on personal or family issues versus college programs. However, for every one th grade students, the relative log odds of them selecting their chosen college based on personal or fami ly issues versus college programs decreased by 0.184. A significant relationship
130 their decision to choose a college based on other reasons not listed versus college prog rams (Table 4 11). Logistic Regression: College Enrollment The fifth question used a linear multivariate regression model to examine if there were significant probabilities between Latina/o students enrolling in college and any of the independent variabl concluded that 63% of Latina/o students enrolled in college within the first two years after their high school graduation and 36% did not. Of the six blocks of variables entered into the regres sion model, only the fourth and fifth block of variables had significant differences (Table 4 12). The fourth and fifth block examined the supply of resources students had access to and their expected college costs. The variables included in these two bloc ks were family income, number of schools applied to, type of college, and when students planned to enroll in college. In particular, the number of college applications students applied to was statistically roll in college or not. For every one unit increase in colleges applied to, the log odds of students enrolling in college increased by 0.422. As for the type of college students planned to attend, this variable was found decision to enroll in college or not. For every one unit change in college type, the log odds of students enrolling in college increased by 0.365. Thus, Latina/o students who planned to enroll in a vocational school versus a four year college were more lik ely to enroll in postsecondary school. significant difference between one of the two variables included: type of high school and degree of urbanization of the school ( Table 4 12). The re was a statistically significant (p
131 private or Catholic school were more likely to attend college. For every one unit change in high school type, the log odds of Latina /o students enrolling in college after high significant variable found in the model. Chapter Summary Chapter 4 presents findings from the advanced analyses. This includes one multilevel ordinal logistic regression, two multilevel logistic regressions, one multilevel multinomial logistic regression, and one logistic regression. The variables significantly associated with 10 th grade Latina/o students college aspirations are; students enrolled in college or not, when all other variables are control led. Students search stage had the most significant variables in comparison to the other models. Finding show that who 10 th grade students went to for college information family th grader, the number of colleges applied, the type of college student planned to attend, and when students planne d to attend. Similar independent variables were found significant in students search stage while in the 12 th r 10 th and 12 th grade student, the number of colleges applied, the type of college student planned to attend, and when students planned to attend.
132 As for students choice stage, the type of college students planned to attend, the desire high school coaches had for them as 10 th th grade, were all found to be statistically significant in examined whether Latina/o students enrolled in college. Only three variables were found significant to this model and they were number of colleges they applied to, the type of college they planned to attend, and the type of high school they attended. The next chapter will expand on the discussion of these statistical findings within the context of the theoretical frameworks used for this study and the extant literature.
133 T able 4 1 Results of m ultilevel ordinal logistic regression model colleg e aspirations Independent Variables Estimate Two tailed p value Demographic Characteristics Gender 0.227 0.016 Family composition 0.031 0.271 Number of in home siblings 0.063 0.359 Social and Cultural Capital 0.07 2 0.243 0.168 0.000 0.043 0.576 0.065 0.135 Social n expectation for 10 th grader after high school a. Mother b. Father c. Friend d. Close relative e. School c ounselor f. Fa vorite t eacher g. Coach 0.044 0.017 0.016 0.031 0.002 0.025 0.010 0.418 0.738 0.591 0.459 0.963 0.597 0.829 Social n expectation for 12 th g rader after high school a. Mother b. Father c. Friend d. Close relative e. School c ounselor f. Favorite t eacher g. Coach 0.037 0.019 0.030 0.036 0.029 0.023 0.002 0.268 0.667 0.341 0.104 0.503 0.604 0.964 Supply of Resources Family income 0.001 0.982 Number of schools applied to 0.061 0.002 ** Expected Costs Type of c ollege 1.364 0.000 *** When s tudent plan ned to enroll in college 0.175 0.003 ** High school & Community Context High school type 0.151 0.060 Degree of urbanization 0.008 0.918
134 T able 4 2 Results of multileve l logistic regres sion model on 10 th grade rs college search stage School Counselor Teacher Coach Independent Variables Estimate Two tailed p value Estimate Two tailed p value Estimate Two tailed p value Demographic Characteristics Gender 0. 081 0. 477 0. 164 0. 152 0 .582 0. 000*** Family composition 0.024 0.005** 0. 023 0. 512 0.027 0. 618 Number of in home siblings 0.093 0.242 0.004 0. 957 0.095 0. 464 Social & Cultural Capital language 0.025 0. 708 0. 013 0 .845 0. 060 0. 604 Engl ish 0.040 0. 505 0.084 0.210 0.0 17 0. 847 status 0.100 0. 275 0.029 0. 767 0.099 0. 495 obtainment 0.024 0. 614 0.006 0. 904 0.055 0.4 98 Social n expectation for 10 th grader a. Mother b. Father c. Friend d. Close r elative e. Counselor f. Teacher g. Coach 0.048 0.034 0.007 0.00 1 0.070 0.001 0.029 0.425 0.609 0.842 0.979 0.304 0.991 0.652 0.077 0.14 9 0.013 0.0 49 0.004 0. 113 0.016 0. 171 0. 022 0. 704 0. 346 0. 947 0. 064 0. 749 0. 012 0.063 0.021 0.069 0.019 0.06 2 0.193 0. 896 0.457 0.709 0.313 0.863 0.462 0.045* Supply of Resources Family income 0.015 0. 753 0.0 55 0.3 02 0.0 57 0. 431 Number of schools appl ied to 0.0 11 0. 018 ** 0.013 0. 469 0.021 0. 520 Expected Costs Type of c ollege 0.129 0. 66 0 0. 273 0. 129 0. 955 0. 000*** When do you plan to enroll in college 0.110 0.132 0.011 0 .830 0.0 97 0.1 94 High school & Community Context High school type 0.145 0. 100 0.136 0. 184 0.061 0. 712 Degree of urbanization 0.0 37 0. 716 0.001 0.996 0.022 0. 868
135 Table 4 3 Results of multileve l logistic regression model on 1 0 th grade rs college search stage Parent Friends Siblings Independent Variables Estimate Two tailed p value Estimate Two tailed p va lue Estimate Two tailed p value Demographic Characteristics Gender 0. 279 0. 021* 0.357 0.002** 0.051 0.685 Family composition 0.0 13 0.718 0.025 0.485 0.036 0.329 Number of in home siblings 0.066 0. 365 0.054 0.486 0.003 0.971 Social & Cultural Cap ital language 0.101 0. 119 0.119 0.095 0.101 0.169 English 0.064 0. 243 0.072 0.222 0.054 0.415 status 0.221 0. 004** 0.057 0.512 0.121 0.163 obtainment 0. 108 0. 03 4 0.024 0.647 0.038 0.455 Social n expectations for 10 th grader a. Mother b. Father c. Friend d. Close relative e. Counselor f. Teacher g. Coach 0.012 0.073 0.008 0.00 9 0.001 0.016 0.037 0.831 0 .203 0.815 0.847 0.992 0.776 0.372 0.089 0.134 0.017 0.068 0.066 0.034 0.032 0.208 0.057 0.668 0.161 0.167 0.586 0.453 0.040 0.000 0.020 0.099 0.061 0.010 0.014 0.446 0.998 0.584 0.050* 0.430 0.887 0.838 Supply of Resources Family income 0.071 0. 166 0.048 0.385 0.012 0.846 Number of schools ap plied to 0. 024 0 .222 0.01 1 0.553 0.003 0.908 Expected Costs Type of c ollege 0. 240 0. 225 0.326 0.045* 0.529 0.023* When do you plan to enroll in college 0. 033 0.561 0.003 0.964 0.072 0.167 High school & Community Context High school type 0. 116 0. 244 0.081 0.408 0.048 0.633 Degree of urbanization 0. 069 0. 459 0.061 0.527 0.112 0.250
136 Table 4 4 Results of multileve l logistic regression model on 10 th college search stage Relatives College Publications College Representatives Independent Variables Estimate Two tailed p value Es timate Two tailed p value Estimate Two tailed p value Demographic Characteristics Gender 0. 375 0. 002** 0.331 0.00 7 ** 0 .111 0. 415 Family composition 0.017 0. 679 0.036 0. 369 0.017 0. 688 Number of in home siblings 0.080 0.287 0.0 61 0 .411 0. 101 0. 262 Social & Cultural Capital language 0. 060 0. 395 0.002 0. 975 0.098 0. 284 English 0.005 0. 941 0.01 2 0.849 0.199 0.007 ** status 0.044 0.598 0.069 0. 403 0. 004 0. 971 ob tainment 0.000 0. 998 0. 120 0. 014* 0.035 0. 541 Social n expectations for 10 th grader a. Mother b. Father c. Friend d. Close relative e. Counselor f. Teacher g. Coach 0.0 63 0.095 0.084 0.063 0.008 0.009 0.00 7 0. 271 0.101 0.027 0.194 0.891 0.895 0.890 0.086 0.086 0.0 37 0.007 0.001 0.005 0.019 0. 182 0.155 0.372 0.882 0.982 0.932 0.783 0.0 91 0.0 99 0.037 0.018 0.044 0.0 27 0.0 08 0. 214 0. 175 0.404 0.772 0. 557 0. 682 0. 918 Supply of Resources Family income 0.0 03 0.961 0.068 0. 171 0.015 0.8 10 N umber of schools applied to 0.007 0. 750 0.0 57 0. 008** 0.0 37 0. 129 Expected Costs Type of c ollege 0. 398 0. 025 0.220 0. 196 0. 686 0. 000*** When do you plan to enroll in college 0.148 0. 037 0.032 0.576 0.080 0. 224 High school & Community Co ntext High school type 0.131 0.2 55 0.065 0.4 70 0.109 0. 376 Degree of urbanization 0.055 0. 591 0.137 0.131 0.013 0.923
137 Table 4 5 Results of multilevel lo gistic regres sion model on 10 th college search stage College Search Guides None of the above Independent Variables Estimate Two tailed p value Estimate Two tail ed p value Demographic Characteristics Gender 0.2 36 0. 060 0.307 0. 025* Family composition 0.008 0. 840 0.044 0.294 Number of in home siblings 0.017 0. 838 0.032 0.744 Social & Cultural Capital 0.024 0.769 0. 039 0. 629 Paren 0.016 0.822 0.001 0.994 0.111 0.210 0.010 0.921 0.08 8 0.0 67 0.026 0. 710 Social n expectations for 10 th grader a. Mother b. Father c. Friend d. Close relative e. Counselor f. Teacher g. Coach 0.0 19 0.071 0.021 0.016 0.056 0.116 0.057 0. 298 0.488 0.723 0.450 0.551 0.814 0.954 0.019 0.071 0.021 0.016 0.056 0.116 0.057 0. 819 0.293 0.623 0.817 0.454 0.078 0.340 Supply of Resources Family income 0. 023 0. 66 2 0.0 1 4 0.812 Number of schools applied to 0.072 0. 001** 0.087 0.001** Expected Costs Type of c ollege 0. 313 0.225 0.875 0.0 00** When do you plan to enroll in college 0.025 0.561 0. 267 0.000*** High School & Community Context High school t ype 0.028 0. 789 0.056 0. 675 Degree of urbanization 0.078 0.439 0.056 0.630
138 Table 4 6 Results of multilevel logistic regression mo college search stage School Counselor Teach er Parents Independent Variables Estimate Two tailed p value Estimate Two tailed p value Estimate Two tailed p value Demographic Characteristics Gender 0. 172 0. 223 0.059 0. 663 0.046 0. 752 Family composition 0.042 0. 383 0.012 0. 776 0.022 0. 604 Number of siblings 0.020 0.847 0.122 0 .160 0.049 0. 562 Social & Cultural Capital language 0.017 0.857 0.169 0. 058 0.247 0. 001** English 0.038 0. 641 0.093 0 .172 0.060 0.359 status 0.3 26 0. 008 ** 0.0 92 0. 314 0.256 0. 003** obtainment 0.048 0. 513 0.026 0. 603 0. 135 0. 024* Social n expectations for 10 th grader a. Mother b. Father c. Friend d. Close relative e. Counselor f. Teacher g. Coach 0. 108 0.088 0.0 40 0.039 0.087 0.096 0.044 0. 197 0.259 0.520 0.572 0.321 0 .277 0.472 0.016 0.080 0.0 27 0.088 0.043 0.042 0.046 0. 843 0.351 0.539 0.110 0.603 0.492 0.422 0.0 70 0.032 0.125 0.102 0.079 0.0 45 0.041 0. 400 0.724 0.003** 0. 109 0. 303 0. 576 0. 471 Social n etwor expectations for 12 th grader a. Mother b. Father c. Friend d. Close relative e. Counselor f. Teacher g. Coach 0.0 15 0.025 0.051 0.002 0.355 0.200 0.0 22 0. 768 0.723 0.290 0.959 0.000*** 0.014* 0.778 0.0 34 0.048 0.047 0.036 0.103 0.027 0.102 0. 411 0.305 0 .255 0.318 0.099 0. 729 0.055 0.07 0 0.069 0.017 0.009 0.151 0.077 0.040 0. 136 0. 243 0.673 0. 786 0. 037 0.277 0. 551 Supply of Resources Family income 0.0 30 0. 668 0.005 0. 921 0.0 73 0. 187 Number of schools applied to 0. 102 0. 000*** 0.0 47 0. 025 0.0 20 0. 388 Expected Costs Type of c ollege 0.334 0. 04 5 0.131 0. 257 0.104 0. 307 When do you plan to enroll in college 0.024 0. 712 0.016 0.774 0.0 05 0. 941 High school & Community Context High school type 0. 297 0. 063 0.030 0. 779 0.181 0. 1 08 U rbanization 0.118 0. 396 0.014 0.890 0.012 0.922
139 Table 4 7 Results of multilevel logistic regression model on 1 2 th grade r s Family Friend College Sources Independent Variables Estimate Two tailed p value Estimate Two tailed p value Esti mate Two tailed p value Demographic Characteristics Gender 0. 097 0. 419 0.263 0. 030 0.577 0. 003** Family composition 0.009 0. 832 0.016 0. 698 0.009 0. 843 Number of siblings 0. 041 0 .607 0. 070 0 .420 0.100 0. 317 Social & Cultural Capital native language 0.053 0 .471 0.1 01 0. 153 0. 057 0. 549 English 0.0 71 0. 304 0.055 0. 278 0.0 52 0 .495 status 0. 291 0. 001** 0.140 0. 194 0.115 0. 273 obtainment 0.0 01 0. 992 0.02 3 0. 700 0. 049 0. 498 Social n expectations for 10 th grader a. Mother b. Father c. Friend d. Close relative e. Counselor f. Teacher g. Coach 0.040 0.057 0.046 0.023 0.0 72 0.069 0.041 0. 669 0.509 0.168 0.688 0.293 0 .342 0.579 0.001 0.04 0 0.0 11 0.074 0.040 0.159 0.096 0. 994 0.567 0.804 0.177 0.597 0.039 0.160 0.0 02 0.058 0.052 0.052 0.049 0.140 0.007 0. 982 0.454 0.332 0. 461 0. 562 0. 113 0. 926 Social n expectations for 12 th grader a. Mother b. Father c. Friend d. Close relative e. Counselor f. Teacher g. Coach 0 .0 09 0.017 0.049 0.032 0.090 0.013 0.063 0. 826 0.715 0.167 0.326 0.242 0.781 0.166 0.072 0.0 23 0.007 0.046 0.123 0.061 0.077 0. 141 0.758 0.853 0.187 0.146 0. 385 0.188 0.011 0.009 0.006 0.090 0.238 0.236 0.012 0. 894 0. 905 0.899 0. 047 0. 001** 0.000*** 0. 872 Supply of Resources Family income 0.0 44 0. 420 0. 018 0. 739 0.0 58 0. 343 Number of schools applied to 0. 020 0. 310 0.0 28 0. 313 0. 179 0. 000*** Expected Costs Type of c ollege 0. 197 0. 061 0.219 0. 053 0. 435 0. 000*** When do you plan to enroll in college 0.048 0. 491 0. 094 0 .107 0.030 0. 677 High school & Community Context High school type 0. 069 0. 437 0.125 0. 255 0.1 12 0. 391 Degree of urbanization 0.110 0. 290 0.028 0. 778 0.171 0.193
140 Table 4 8 Results of multilevel logistic regression mo del college search stage Libraries None of the above Independent Variables Estimate Two tailed p value Estimate Two tailed p value Demographic Characteristics Gend er 0. 178 0. 250 0.3 41 0. 207 Family composition 0.00 4 0. 915 0.005 0. 944 Number of in home siblings 0.0 05 0. 972 0.057 0.7 15 Social & Cultural Capital 0.0 41 0 .654 0.0 54 0.6 69 0.011 0. 900 0 .064 0.507 0.009 0. 940 0.0 51 0. 725 0.021 0. 744 0.02 0 0. 831 Social n expectations for 10 th grader a. Mother b. Father c. Friend d. Close relative e. Counselor f. Teacher g. Coach 0.0 50 0.109 0.038 0.104 0.033 0. 094 0.082 0. 451 0.133 0.387 0.065 0.711 0.217 0.224 0.004 0.007 0.0 30 0.015 0.0 42 0. 022 0.05 6 0. 972 0.929 0.658 0.886 0.750 0.865 0.660 Social n expectations for 12 th grader a. Mother b. Father c. Friend d. Close relative e. Counselor f. Teac her g. Coach 0.01 7 0.029 0.076 0.051 0.023 0.090 0.042 0. 728 0.645 0.097 0.277 0.720 0.230 0.489 0.0 0 8 0.134 0.055 0.029 0.371 0.372 0. 1 19 0. 945 0.287 0.430 0.657 0.000*** 0.000*** 0.150 Supply of Resources Family income 0.039 0.6 02 0.0 54 0. 563 Number of schools applied to 0.025 0. 367 0. 160 0.000 ** Expected Costs Type of c ollege 0. 222 0.2 41 0. 336 0. 058 When do you plan to enroll in college 0.012 0.883 0. 054 0. 557 High School & Community Context High school type 0.070 0. 549 0. 261 0. 174 Degree of urbanization 0.005 0.969 0.277 0. 072
141 Table 4 9 Results of multilevel multinomial logistic regression model o n 10th grade rs Compared to College Programs Independent Variables Estimate Two tailed p value Estimate Two tailed p vale Demographic Characteristics Gender 0.010 0.956 0.160 0.527 Family composition 0.011 0.833 0.025 0.736 Number of in home siblings 0.021 0.846 0.083 0.538 Cultural & Social Capital 0.041 0.704 0.036 0.759 0.012 0.886 0.024 0.825 0.056 0.686 0.074 0.641 0.042 0.603 0.140 0.143 Social n expectations for 10 th grader a. Mother b. Father c. Friend d. Close relative e. Counselor f. Teacher g. Coach 0.107 0. 136 0.0 78 0.069 0. 056 0.130 0. 1 13 0.188 0.119 0.164 0.453 0.682 0.166 0.235 0. 279 0.228 0.000 0.078 0.0 72 0.035 0.119 0.065 0.074 0.997 0.554 0.713 0.792 0 .207 Social n expectations for 12 th grader a. Mother b. Father c. Friend d. Close relative e. Counselor f. Teacher g. Coach 0.0 67 0.169 0.0 24 0.0 67 0.0 80 0. 033 0.0 55 0.244 0.055 0.656 0.129 0.317 0.653 0.505 0.046 0. 068 0.041 0.060 0.021 0.180 0.164 0. 528 0.512 0.542 0.271 0.828 0. 056 0.08 5 Supply of resources Family income 0. 110 0. 132 0.019 0.839 Number of schools applied to 0.005 0. 887 0.048 0.213 Expected Costs Type of c ollege 0. 320 0 .037* 0.354 0.102 When do you plan to enroll in college 0.005 0. 950 0.033 0.758 High School & Community Context High school type 0. 057 0. 690 0.085 0.588 Degree of urbanization 0.136 0. 335 0.142 0.448
142 Table 4 10 Results of multilevel multinomial logistic regression model o n cho ice stage Compared to College Programs Independent Variables Estima te Two tailed p value Estimate Two tailed p vale Demographic Characteristics Gender 0.051 0.843 0. 315 0. 075 Family composition 0.017 0.808 0 .013 0.789 Number of in home siblings 0.169 0.274 0.225 0.067 Cultural & Social Capital ative language 0.063 0.673 0.014 0.883 0.108 0.262 0.02 6 0 .745 0.026 0.876 0.018 0.874 0.0 16 0 .873 0.019 0.776 Social n expectations for 10 th gr ader h. Mother i. Father j. Friend k. Close relative l. Counselor m. Teacher n. Coach 0.051 0.029 0.140 0. 101 0.100 0.024 0.171 0.711 0.824 0.081 0.256 0.576 0.846 0.154 0. 160 0.127 0.021 0.0 11 0.175 0.157 0.170 0.268 0.239 0.741 0.870 0.051 0 .055 0.012 Social n e expectations for 12 th grader h. Mother i. Father j. Friend k. Close relative l. Counselor m. Teacher n. Coach 0.026 0.061 0.006 0.054 0.008 0.081 0.087 0.691 0.519 0.941 0.345 0.936 0.522 0.437 0.100 0. 150 0.001 0.0 07 0.055 0.18 4 0.66 0.067 0.112 0.981 0.8 50 0.350 0.007 ** 0.343 Supply of resources Family income 0.014 0.894 0.019 0.793 Number of schools applied to 0.054 0. 147 0.0 60 0 .075 Expected Costs Type of c ollege 0. 596 0 .015 0. 107 0 .386 When do you plan to enroll in college 0.069 0.547 0.009 0.916 High School & Community Context High school type 0.0 85 0. 653 0.187 0. 170 Degree of urbanization 0. 002 0 .992 0. 030 0. 847
14 3 Table 4 11 Results of multilevel multinomial logistic regression model o n 1 2th grade rs choice college Independent Variables Estimate Two tailed p value Demographic Characte ristics Gender 0.0 40 0 .834 Family composition 0.0 57 0 .404 Number of in home siblings 0.0 1 7 0.912 Cultural & Social Capital 0.021 0.860 0.036 0.791 0.008 0.914 0.021 0.822 Social n expectations for 10 th grader o. Mother p. Father q. Friend r. Close relative s. Counselor t. Teacher u. Coach 0. 023 0.023 0.029 0.0 31 0.189 0.1 79 0.0 13 0.881 0.882 0.667 0.728 0.123 0.268 0.935 Social n etwor expectations for 12 th grader o. Mother p. Father q. Friend r. Close relative s. Counselor t. Teacher u. Coach 0.0 25 0.002 0. 173 0.015 0.0 3 0 0. 166 0.172 0.737 0.989 0.016 0.774 0.803 0.074 0.060 Supply of resources Family income 0.021 0.831 Number of schools applied to 0.018 0. 677 Expected Costs Type of c ollege 0. 059 0 .711 When do you plan to enroll in college 0.011 0.916 High School & Community Context High school type 0.183 0. 375 Degree of urbanization 0.061 0.727
144 Table 4 12 Results of l college enrollment College Enrollment Status Independent Variables Estimate Two tailed p value Demographic Characteristics Gender 0.2 82 0. 054 Family composition 0.0 73 0. 137 Number of in home siblings 0.045 0. 645 Cultural & Social Capital 0.04 2 0. 665 0.133 0.082 0.131 0 .253 onal obtainment 0.087 0. 150 Social n expectations for 10 th grader a. Mother b. Father c. Friend d. Close relative e. Counselor f. Teacher g. Coach 0.0 13 0.096 0.0 48 0. 032 0.0 93 0.007 0.042 0. 838 0.151 0.303 0. 603 0.374 0.938 0.570 Social n expectatio ns for 12 th grader a. Mother b. Father c. Friend d. Close relative e. Counselor f. Teacher g. Coach 0.00 7 0.0 05 0.0 28 0.023 0.062 0.018 0.004 0. 876 0.933 0.542 0.577 0.356 0.797 0.948 Supply of resources Family income 0.0 52 0. 401 Number of schools applied to 0. 422 0. 000*** Expected Costs Type of c ollege 0. 365 0.002** When do you plan to enroll in college 0.106 0.085 High school & Community Context High school type 0.712 0. 000*** Degree of urbanization 0.152 0.217
145 CHAPTER 5 DISCUSSION OF FINDINGS choice, which guided t his study and will be further discussed. To achieve this, Chapter 5 will review the purpose of this study and the research questions that guided this study. Next, the findings will be discussed within the context of the extant research on Latina/o students research will be further elaborated within C hapter 6 These findings will be presented based on the layers and sections of the conceptual frameworks used in this study. This study further develops the current literature by examining potential differences that exist within Latina/o students population and recognizing they are not a monolithic group but a heterogeneous population. Furthermore, this study provides insight into decision to enroll in college, specifically whether certain social networks influenced their decisions. Overall, this study contributes to the literature by identifying factors that are area that has a dearth of empirical evidence on this increasingly heterogeneous group. Purpose of Study Revisited First generation students continue to see a n increase in their access to postsecondary institutions (Choy, 2001; Strayhorn, 2006). It is estimated that 30% of all college students are considered first generation (Strayhorn, 2006). Despite increased access to postsecondary institutions, first genera tion students continue to differ from their peers when navigating the college choice process, especially in relation to college
146 enrollment (Inkelas & McCarron, 2006; Saenz, et al., 2007). Specifically, first generation students typically have lower educati onal aspirations than non first generation (Lohfink degree aspirations since they often influence whether or not a student will actually pursue a college education. After all, aspirations are typically one of the strongest students impacts their colleg e choice process and their likelihood to enroll in college. There are four research questions guiding this study: Does the expectations social networks have for Latina/o high school students, along irations? Does the expectations social networks have for Latina/o high school students, along entrance information in both the 10 th and 12 th grade? Does the expectations soci al networks have for Latina/o high school students, along with other key variables, influence the primary reason students selected their chosen college? Does the expectations social networks have for Latina/o high school students, along with other key var iables, influence their decision to enroll in college? In order to examine these research questions, the Educational Longitudinal Study of 2002 was selected as the dataset for this study because it provided the most current and robust dataset available to Conducting descriptive and logistic regression analyses of the dataset provides information describing potential characteristic differences of Latina/o students associated with how they navigate the college ch oice process and college enrollment. The following section will discuss the findings of this empirical study based on the theoretical frameworks employed throughout the study.
147 Summary of Research Findings The following sections will present the results f rom the data analysis within the decision to enroll in college or not. To facilitate the discussion of the results, the findings will be presented based on the adapted version model of college choice, as previously discussed. In particular, demographic characteristics will be discussed first, followed by cultural and social capital, supply of resources, expected costs, and finally school co (2006) proposed conceptual model of college choice provides a comprehensive lens to extant literature due to its robust inclusion of v ariables from both economic and sociologic approaches. Demographic Characteristics In order to examine how Latina/o students navigate the college choice process, three demographic variables were selected for this study: gender, family composition and nu mber of siblings living at home. The first demographic variable, gender, was not found significant across all models. However, gender was a significant predictor of femal e students to have similar levels of degree attainment (Inkelas & McCarron, 2006). While not all models reported significant findings comparing 10 th grade Latino male and female students, there were some gender differences found significant and are importa nt to examine. Primarily, Latina/o male students (36%) were less likely than Latina females
148 school diploma (Table 5 10). Current research has found that male students, regardless of ethnicity, exhibit lower educational expectations than their female peers (Lowman & Elliott, 2010; Mau & Bikos, 2000; Park, et al., 2011). However, earlier research indicated that males typically had higher college aspirations than females ( Wilson & Wilson, T able 5 were not that much higher (22% vs. 18%) than Latino males. Regardless of the conflicting findings in the literature, it is apparent that gender does play a role in Latina/o explore at what age and grade level Latina/o students sta rt considering a high school diploma as a viable final option for their educational aspirations. Secondly, 10 th grade Latina/o students were more likely to go to their parents (47%) for college entrance information than any other source; followed by their guidance counselor (45%) and friends (43%). However, from these three top sources gender was only found significant among 10 th grade students who seek college information from their parents. Gender was also found significant among the following college se arch sources: coaches, friends, relatives, college publications, and none of the above. According to the findings, 10 th Latino males (62%) were more likely to go to their high school coach (Table 5 6) for college information versus 10 th Latina females (47% ) who were more likely to go to their parents (37% vs. 31% males), friends (69% vs. 61% males), relatives (76% vs. 69% males), and college publications (80% vs. 74% males). Since males are more involved playing sports throughout high school compared to fem ales, it was no surprise they were more likely to seek college information from
149 their coaches. Latino males were also more likely (49%) than Latinas (41%) not to seek college information from any of these sources (Table 5 8). But by the 12 th grade only Lat inas were more likely to seek college entrance information from either their friends (72% vs. 67% males) or college sources (93% vs. 89% males) than males (Table 5 10). Although previous studies have found contrasting findings on the significance of gende role of gender on college choice decisions. Especially, as females increase their enrollment rate in postsecondary institution and outnumber their male peers (Snyder & Dillo w, 2011). This study indicated that even as early as the 10 th grade Latino males are already exhibiting lower educational aspirations than their female peers (Table 5 1). Therefore, further examination is necessary to understand what factors are contributi ng gender and how they navigate the college choice process (Park, et al., 2011). Family composition and number of siblings living at home are two additional variables exa mined in relations to the college choice process of Latina/o students. Family composition was only found significant across one model; 10 th graders college search stage. Latina/o students who came from two parent households (57%) versus a single parent hou sehold (55%) were less likely to seek college information from their school counselor as 10 th graders. In regards to the number of siblings living at home, this variable was not found significant. The majority (47%) of Latina/o students reported to have le ss than two siblings living in the same household as them. There is the possibility that Latina/o students in this sample shared similar family compositions and number of siblings living at home compared to students of difference racial groups or
150 generatio ns. Nevertheless, more exploration is necessary to determine if these proposed conceptual model of colleg e choice, as well as the extant literature. Previous college choice process. Therefore, it is feasible to say other factors examined in this study might have limited the influence of family composition and number of siblings living at home. Additionally, variables that were not available in ELS:02 might have significant level of their variables. Cul tural Capital The next set of variables examined in this study were centered around the idea due to the inclusion of sociological factors. This study selected four vari ables to key component in the college choice process. After all, many students rel y on others to gain knowledge about future opportunities. Oftentimes these influences come directly from their parents (McDonough, 1997; Tierney & Venegas, 2006). Unfortunately, first generation students are often placed at a further disadvantage since man y lack the necessary cultural capital resources (Terenzini, et al., 1996; Walpole, 2003).
151 Perna 2008). Findings from this study support these claims. For example, students whose parents only earned a high school diploma or less, compared to students with college educated parents, were more likely to obtain college entrance information from their pa rents in both the 10 th ( Table 5 6) and 12 th grade (Table 5 7), as well as from college publications ( Table 5 9). However, this does not imply students are more likely to seek these sources for college entrance information over other sources. But rather tha was not found significant in the other college choice models. It is well documented th at students who identified as first generation are more likely to come from homes where English is not the primary language (Bui, 2002). degree aspirations, regardless o th likely to seek college infor mation from their parents as 12 th graders than students whose native language was S panish or another language ( Table 5 9). However, there was no stage or enrollment sta ge. Despite not being found significant in most models, this variable should not be excluded from future studies because it does contribute to the
152 native language was not process, other research has found that when English is the second language at home, educational aspirations may be lower (Behnke, Piercy, & Diversi, 2004). In addition to examining the influe college choice process, this study sought to investigate if there was a significant choice process. This study found that Latina/o students whose parents were native compared to students whose parents were not fluent English speakers (36%). Overall, students whose parents did not speak English or w ere partially fluent indicated a higher parents were fluent or native English speakers (Table 5 2). Despite parents limited proficiency in English, their expectations or p ersonal struggles may have had a bigger As for the search stage, there was only a significant relationship found between secondary information from college representatives as 10 th graders. In particular, students whose parents were fluent (93%) or native English (94%) speakers were more likely to seek college information from college representatives compared to students whos e parents were not fluent English speakers ( T able 5 Salazar, 2004); therefore, scant information is available in the ext ant literature. This
153 variable may have not been as significant as others as a result to the many independent variables included in each model. This study also examined whether a significant difference existed between ess and their generational status. This variable was only found significant in the college search stage models. For the first search model, 10 th grade immigrant students were more likely (31%) to seek college information from their parents than second gene ration (23%) Latina/o students (Table 5 6). However, by the 12 th grade second generation Latina/o students (45%) were more likely to seek college information from their parents than immigrant students (33%) (Table 5 9). On the other hand, immigrant student s in 12 th grade were more likely to seek college information from their school counselors (60%) and family members (58%) than second generation Latina/o students (Table 5 9). col lege choice process in relations to their generational status. The most common and degree aspirations. According to the literature, immigrant youth move upward in the educational ladder compared to their parents but this trend reverses by the second not found significant across the other models in particular, the predisposition, choi ce, generational status is influential in the college choice process when controlling for additional variables or eliminating some of the variables used in this study.
154 Overall, th is study suggests that particular aspects of cultural capital are variables. Prior researc choice process (Lowman & Elliott, 2010; McDonough, 1997; Tierney & Venegas, 2006), but oftentimes researchers treat Latina/o students as a homogeneous group. The findings in this study emp hasize the need for Latina/o students to be directed to the proper sources where they can obtain accurate college information. As well as, be encouraged early on to participate in college discussions therefore, their chances of navigating the college choic e process can be enhanced. Social Capital proposed conceptual college choice model. This particular study examined social ompletion. These two independent variables BYS68 and F1S44 were significant in the second and third multilevel model which measures the various sources students went to for college information while in the 10 th and 12 th grade, respectively. As well as, the multinominal model which measures the main reasons students decided to attend their chosen college. This section will start off with those social networks that were the most significant towards Latina/o students search stage and so on. In general, L atina/o students were more likely to seek college entrance information from various sources as 12 th graders than as 10 th graders. For example, 62% of students went to their school counselor for college information as 12 th graders versus 45% who went to the ir guidance counselors as 10 th graders. Overall, the
155 expectations school counselors had for 12 th grade students was a significant predictor to the type of sources students were most likely to seek as seniors. Latina/o students whose school counselor desire d for them to do as they wanted after high school were more likely to seek college information from them (90%) than students whose counselor desired for them to attend college (60%). Similar results were found among school ts and their decision to seek college i nformation from their parents ( Table 5 9) and other college sources (Table 5 10). college were more likely (61%) to seek college informati on from none of the above whatever they wanted to after high school (Table 5 10). Many studies have proven that ege choice stages, in particular first generation students (McDonough, 2004, Bryan et al., 2011). Although, the findings do suggest that information from certain sources. Similar to guidance counselors, more students went to their teachers for college entrance information as the student was only found significant in the search and choice stage. More specifically, college informat ion from their school counselor, friends, various college sources, and
156 students, Latina more likely to seek college information from their 12 th school counselor (60%) and and 71%, respectivel y. On the other hand, Latina/o students whose 10 th desire for them to attend college were less likely to obtain college information from their please (82%) aft er high school (Table 5 10). The literature provides contradicting findings on the role teachers have on aw potential in them and challenged them to persevere (Butner et al., 2001; Rooney, 2008). Unfortunately, many students come across teachers and counselors who send adverse messages or hold lower expectations for them, which often delimits their college as pirations (Butner in both their search and choice stage. Teachers who desire for their students to attend college impacted the type of resources students were mos t likely to seek for college information. Teachers who desire for students to attend college were found to be on personal or family issues versus other reasons (Tabl e 5 11). More research studies their level of significance and not only in the predisposition stage.
157 Moreover, a suitable amount of Latina/o students obtained college in formation from their friends as 10 th graders (43%) and as 12 th had for Latina/o students were only found significant in the search and choice stage. ision to seek college information from their relatives as sophomore (Table 5 7) and from their parents as seniors ( Table 5 9). As well as, influence their decision to select a given college (Table 5 obtain a college degree were less likely to seek college information from relatives (69%) than Latina/os s likely to seek college information from their parents (33%) as seniors (Table 5 9). In terms of the select their chosen college based on other reason not listed ( Table 5 11). aspirations of attending college (Contreras (2005) found that peers negatively impacted the college aspirations of Latina/o studen ts that students were often encouraged by their friends to attend college, but usually when ege aspirations (Contreras Godfrey, 2009; DiMaggio & Mohr, 1985; Gomez, 2005). More students should have access to peer group participation who value higher educatio n and frequently converse about their future education (DiMaggio & Mohr,
158 1985). For this reason, more research is needed to examine the level of influence As for close relatives, 30% of Latina/o s tudents went to relatives for college entrance information. The variables siblings and close relatives were combined and sus doing something else ( Table 5 7) were less likely to seek college information from their siblings as 10 th graders. Then again, 10 th grade may be too early for Latina/o students to start gathering college information from various sources. Nonetheless, L for them to attend college versus doing something else were less likely to seek college information from college sources (Table 5 10). Although very limited research is done on the influence family members have on stu doubt they play a critical role in all three stages. The findings may have been drastically different had students of various races been incorporated. 10 th grade search stage (Table 5 6) and their 12 th grade choice stage were coaches (Table 5 11). Latina/o college information from them (62%) than students whose d esire for them to get marry (43%) or do whatever they wanted to (38%). In terms of the choice stage, marry, were less likely to select their chosen college based on personal or family issues ( Table 5 college choice process. Thus, more research should examine the level of influence high
159 choice stage and attend college, especially for male students. Contrary to other sources, the percent of Latina/o students who reported to go to their parents for college entrance information in the 10 th grade decreased by the 12 th grade, 47% versus 39% to seek college information from other sources as 12 th graders. Earlier research (Chapman, 1981). However, othe r research highlights that constituents, such as peers al., 1989; Smith & Fleming, 2006). A significant difference was only found between the their 10 th grade students and their decision to seek college information from their teachers ( T able 5 6). Further research should examine the especially among single family households. were the most significant across Latina/o students search stage. This does not mean choice process. Although, it is possible that other variables may have confound some elements of this independent variable (school counselors) from being significant across the other two multilevel logistic regression models which examine stud stage both as 10 th and 12 th n
160 eliminating some of variables or when running different models. The extant literature on example, recent research has suggested that parents, guidance counselors, teachers, (McDonough, 2004; Bryan et al., 2001; Contreras Godfrey, 2009; Rooney, 2008). Other researchers have indicated contradictory findings (L.X. Perez, 1999; Butner et al., 2001; Gome z, 2005) which is why more research should look at the influence of social Supply of Resources social capital in the college choice p rocess and suggests that key variables such as income and information about college were integral to studying college choice. Earlier college choice models have been criticize for examining the college choice process from the perspective of only high incom e families (Hurtado, et al., 1997). Therefore, this study selected two variables to investigate supply of resources: family income levels and number of college applications submitted. Collectively, these two variables created important aspects under supply of resources. their family income levels (Fitzgerald, 2004; Trusty, Robinson, Plata, & Ng, 2000). Some research suggests that low socioeconomic status is one of the strongest p redictors of lower educational expectations (Hanson, 1994). However, this study does not support or discredit these claims since there was no significant difference found between family income levels and any of the multilevel models. Then again, the variab le number of schools applied to may have reduced or canal out the level of significance family
161 submitted (M.T. Hurtado, 1997; L. Perna, 2000). Thus, Latina/o students who come from middle or high income households were probably more likely to apply to more colleges compared to low Contra ry to family income, the number of college applications Latina/o students applied to were found significant in all of the multilevel regression models expect the multinomial model which encompassed the main reason students selected their chosen college. Fi college applications they applied to. Latina/o students who applied to more than three colleges were more likely to aspire to obtain a college degree and enroll in college t han students who applied to only one college (Table 5 3). However, there was no drastic difference among students who applied to three schools (37%) versus one school (36%) 3). Previous r esearch (M.T. Hurtado, 1997) has found family income to be significant to the number findings do not support this claim the number of schools students apply to are a ke y level of significance is small. Moreover, the number of college applicants Latina/o students applied to during s 10 th and 12 th graders. For instance, Latina/o students who applied to more than three colleges were more likely to seek college entrance information fro m their guidance counselor ( Table 5 6), college publications ( T able 5 7), and college source guides
162 (T able 5 8) as 10 th graders. On the contrary, students who applied to less than two colleges versus students who applied to five schools, did not seek college information from any of the above sources (Table 5 8). Similar findings were found among the number of college applications students applied to and who they went to for college information as 12 th graders, in particular school counselors and college sources. Once again, Latina/o students who applied to more than three schools were more likely to obtain college information from their school counselor (69%) in comparison to students who only applied to one school (60%). Regardless of how many college applications they submitted, more students pursue college information from these sources as 12 th graders t han as 10 th graders. Latina/o students who applied to more than one college were more likely to seek college information from their teacher ( T able 5 9) and college sources (Table 5 10). As for students who applied to one school (61%) versus four (49%) were more likely not to seek college information from any of the above sources (Table 5 10). Thus, the number of schools students submitted were found to be statistically significant to the sources students seek the most for college. There was no significant difference found between the number of applications students applied to and the choice stage. As for college enrollment, there was a significant difference found between the number of college applications students applied to and their decision to enroll in college. For example, students who applied to only one college were less (27%) likely to enroll in college than Latina/o students who applied to applied to the more likely they were to aspire for a college degree, seek college information from school counselors, teachers, and college sources, as well as enroll in
163 college. Nonetheless, additional research should investigate the degree of significances applying for college app inclusion or elimination of other variables or statistical models may have enhance the degree of significances college applications have on the college choice process. Expected Cost (2006) model also incorporates the expected cost of college attendance. The two variables examine under this block were the type of institution students planned to enroll in and when they planned to enroll in college, if at all. Collectively, these two var iables address the expected cost and its role to the college choice process of Latina/o students. The only independent variable found statistically significant among all the advanced analyses models was the type of college students anticipated enrolling in Based off the multilevel ordinal logistic regression model, a significant difference existed between the type of postsecondary institutions students planned to attend and their college aspirations. Latina/o students who planned to attend a community coll ege planned to attend a four year college (Table 5 4). Unfortunately, many studies do not take into consideration the type of institutions, two year versus four year, stu dents are more likely to aspire to attend (Anderson, 2008; Carreras, 1998; Gomez, 2005; Kao & Tienda, 1998; P.A.Perez, 2007; Wolf, 2007). Although it may seem reasonable for should examine at what age and grade level students start to decide which type of college they plan to enroll in as a practicable option for their educational aspirations. Moreover, 10 th grade Latina/o students who planned to attend a four year college ve rsus a vocational school, were more likely to seek college entrance information from
164 their coaches (62% vs. 19%), siblings (48% vs. 24%), friends (61% vs. 45%), close relatives (69% vs. 50%), and college representatives (88% vs. 66%). Ironically, Latina/o students who planned to enroll in a four year college versus a vocational school were more likely not to seek college information from any of the above sources (Table 5 8). Similar findings were found among 12 th grade students who aspired to attend a vocat ional school versus a four year college. In particular, students who planned to enroll in a four year college were less likely to seek college information from school counselors (Table 5 9) and college sources (Table 5 10) than students who aspired to enro ll in a vocational school. In regards to the choice stage, Latina/o students who aspired to attend a reputation and location compared to students who planned to attend a four ye ar institution ( T able 5 11). A possible explanation for these findings may have to do with students' certainty of attending the best vocational school in their field of interest. According to the literature, Latina/o students are more likely to attend a co llege based Fry, 2006). Unfortunately, ELS:02 does not specify whether the variable location takes ited research research studies. However, the type of institution students aspired to attend is critical to their college choice process and requires further examination. F urthermore, when looking at the binary logistic regression model, the type of college Latina/o students planned to attend was a significant predictor of college
165 enrollment (Table 5 12). Unfortunately, whether students who aspired to attend a vocational sch ool actually enrolled in a four year, two year, or vocational school is left unknown because the dependent variable, F2B07, only measured college enrollment. However, there was no clarification in the type of institution students actually enrolled in. Rese archers should further examine how the type of college students plan to enroll in affect their college choice process because students chances of earning a college degree decreases if they start off at a two year institution verses a four year college (Fry 2006). Despite the degree of significances, the type of postsecondary institution a student plans to attend is influential in their decision to choose a particular college and enroll in college. The second variable examined under expected costs was when planned to enroll in college, if at all. This variable was found significant in only two models. Latina/o students who planned to enroll in college after high school were more likely to have higher degree aspirations than students who planned to delay college enrollment after a year or more (Table 5 5). Additionally, 10 th grade Latina/o students who anticipated postponing college were more likely to seek college admissions information from their relatives than students who planned to enroll after high school (Table 5 7). Similarly, students who anticipated enrolling in college after high school were less likely not to seek college information from any of the above sources compared to students who planned to postpone their education ( T able 5 8). Pe rhaps decision to postpone college.
166 Almost half of the students in this sample were first generation students. According to the literature, first generation students oft en struggle more when navigating the college choice process since information that is typically conveyed from parents to the student is often unavailable to students whose parents do not have first hand knowledge of the college choice process (Inkelas & Mc Carron, 2006). Furthermore, this study did support the notion that students who anticipate enrolling at a four year institution were significantly more likely to seek college information from various sources. Despite some students that come from disadvanta ged situations and have parents that lack experience in the college choice process, many still overcome these disadvantages by seeking proper college information from vital sources. School Context The last block of variables investigated in this study wer e based from layer two of students attended. Collectively, these two variables address the school context of the theoretical framework guiding this study and there role on the college choice process of Latina/o students. ocess and enrollment. Nonetheless, this does rather this variable should be examined independently. However, a more nuanced look at the socioeconomic levels of schools indicate that this factor, in addition to school year institution (Engberg &
167 measure of the socioeconomic status of schools because Harwell and LeBeau (2010) provide better insight into whe In addition to examining school urbanization, this study also investigated whether the type of school (public, private, or Catho lic) students attended was statistically significant to their predisposition, search, choice, and enrollment stage. Examining the school context is critical when studying the college choice process of Latina/o students. ignificant difference was found between the type of high school Latina/o students attend and their decision to attend college or not (Table 5 12). While it is assumed that many students attend less funded, less challenging, and lower income schools, (Choy, 2001; Hamrick & Stage, 2004) this is not the case for all Latina/o students. As more Latina/o students navigate the college choice process and enter postsecondary education, it is imperative to see how the role of schools influence e process. This study only incorporated two school contexts variables due to the limitation of the dataset used for this study, but more robust and comprehensive school context variables should be considered in future studies. Despite only using two variab les to measure school context, this study still found support for influential in the college choice process of Latina/o students. Final Reflection of Findings This study s Latina/o students after high school influence the college choice process of a national
168 representative sample of Latina/o students. The previous sections discussed the findings within the cont ext of the theoretical frameworks guiding this study. There are overall themes that span across the five models that warrant further discussion. These themes will be addressed in this section. While the main focus of this study was not found to be signific antly related to all the models, these findings do not dismiss the th and 12 th grade search stage, as we information from various sources. As supported by the literature, institutional gatekeepers can ei covert or overt message they send to students (Butner et al., 2001; Rooney, 2008). ng college information. For example, first generation Latina/o students were more likely to seek college knowledge from their own parents, despite their parents lack of experience in higher education. As well as from college publications than students who repeatedly noted to be lower for first generation students (Choy, 2001; McDonough, 1997; Nunez & Cuccaro Alamin, 1998; P ascarella, et al., 2004). However, there was no degree aspirations. Thus, parental education must be further examined to see what particular aspects influence first generati
169 college choice model. For example, Latina/o students whose parents were native English speakers were more likely to aspire for a college degree than students whose parents did not speak English. By the 12 th grade, immigrant students were more likely to seek college information from their school counselor and relatives than second generation students. The only difference between 10 th and 12 th grade was that immigrant students were less likely to seek college information from their parents by the 12 th grade compared to second generation students. All encompass ing, these findings illustrate how cultural capital, in the sense of family and language, is important to the college choice process. In agreement to previous studies, Latina/o students are a heterogeneous population and their within group diffe rences have had significant influences on how they navigate the college choice process (Phinney, et al., 2005). A final finding from this study that contributes to the extant literature is the (Bergerson, 2009), its inclusion of both economical and sociological fa ctors was found suitable for this study. For example, the variables under supply of resources and expected costs were the most significant among all five regression models. Therefore, an emphasis for Latina/o students to apply to multiple colleges, as well as informing low income or first generation students about college application waivers their eligible for, is much needed.
170 Moreover, under expected costs the type of college Latina/o students intended to enroll in was significant across all the models. I n particular, Latina/o students who aspired to enroll in a vocational school were more likely to choose a college based on reputation and location. They were also more likely to enroll in college than students who planned to attend a four year institution. In agreement to other studies (Fry, 2006), the type of college students intend to enroll in influence their probability of obtaining a should encourage students to enroll in four year institutions verses two year institutions. When students planned to enroll in college was a significant predictor of their predisposition and search stage. According to the literature (Horn, Cataldi, & Sikora, 2006), Latina/o students who pos tpone college were less likely to graduate from college with a degree. Therefore, a college going culture is necessary so students can be encouraged and motivated to attend college as traditional college aged students. In addition to reviewing general fin dings from this study, it is important to note those variables included in ELS:02, excluding several variables that were included in odel. Thus, for future studies more robust social model. Due to the complexity of the research questions it was not feasible to this for this study. Another limitati on pertained to the questions 10 th grade students were asked about their degree aspirations and college choice process. Some researchers have argued that Latina/o students do not necessary follow the same college choice process as middle class White studen ts. Therefore, many Latina/o students were probably
171 uncertain of their degree aspirations due to their age and grade level. As a result, these findings make it difficult to compare those students who were unsure about their future plans as sophomores to th ose who were certain about their educational aspirations. Nevertheless, the findings for this study support the relevance of both models multitude of variables selec support findings in the current literature, while simultaneously providing new information to fill current gaps in the re search. This study is able to provide relevant information about potential factors that positively or negatively influence the college choice pathways of Latina/o students and their decision to enroll in postsecondary institutions.
172 Table 5 1. Results for multilevel ord inal logistic regression model on degree aspiration and gender Degree Aspiration s High school or less Associate of Art incomplete degree Advanced degree Male 4% 2% 41% 36% 18% Female 3% 1% 36% 38% 22% Table 5 2. Results for multilevel ordinal logistic regression model on Degree Aspiration s High school or less Associate of Art degree, incomplete Bach degree complete Advanced degree Not fluent 4% 2% 41% 36% 18% Partially fluent 5% 2% 44 % 34 % 16 % Fluent 6% 2% 47 % 32 % 14 % Fluency unknown 7% 3% 50 % 29 % 12 % Native English speaker 8% 3% 52 % 27 % 10 % Table 5 3 Results for multilevel ordinal l ogistic regression model on and number of schools applied to Degree Aspiration s High school or less Associate of Art degree, incomplete degree, complete Advanced degree Applied to one school 4% 2% 41% 3 6% 18% Applied to two schools 4% 2% 39% 36% 19% Applied to three schools 4% 2% 38% 37% 20% Applied to four schools 3% 1% 37% 37% 21% Applied to five schools 3% 1% 36% 38% 22% Table 5 4. Results for multilevel ordinal logistic regression model on stu of college Degree Aspiration s High school or less Associate of Art degree, incomplete degree, complete Advanced degree Four year college 4% 2% 4% 36% 18% Two year college 1% 0% 17% 36% 46% Vocat ional school 0% 0% 5% 18% 77% Table 5 5. Results for multilevel ord inal logistic regression model on when students plan to enroll in college Degree Aspiration s High school or less Associate of Art degree, inco mplete degree, complete Advanced degree Yes, after high school 4% 2 % 41 % 3 6% 18 % Yes, after one year 5% 2 % 44 % 34 % 1 5% Yes, over one year 6% 2 % 47 % 32 % 13 % 7% 3 % 50 % 2 9% 11 % my education 8 % 3 % 52 % 27 % 1 0%
173 Table 5 6. Results for multile college search choice Variable name Exp(B) e /1+e School Counselor Family composition Both parents 1.3 51209 57% One parent & one guardian or tw o guardians 1.32313 56% Mother only 1.295634 56% Father only 1.268709 55% One guardian only 1.242344 55% Parent or g uardian lives with student less than half the time 1.216527 54% Number of schools applied to Applied to one schoo l 1.351209 57% Applied to two school s 1.366155 58% Applied to three school s 1.381265 58% Applied to four school s 1.396543 58% Applied to five school s 1.41199 59% Teacher th grade student Attend college 4.80 6648 83% Get a job 5.578947 85% Attend trade school 6.475333 87% Join the military 7.515744 88% Get marry 8.723321 90% Whatever the student wants to do 10.12492 91% Coach Gender Male 1.601595 62% Female 0.894939 47% Coac th grade student Attend college 1.601595 62% Get a job 1.320486 57% Attend trade school 1.088717 52% Join the military 0.897628 47% Get marry 0.740078 43% Whatever the student wants to do 0.610181 38% Type of co llege student plans to attend Four year college 1.601595 62% Two year college 0.616313 38% Vocational school 0.237165 19% Parents Gender Male 0.444413 31% Female 0.587429 37% Generational Status Immigrant 0.444413 31% First generation 0.359874 26% Second generation 0.291417 23% Parental Education High school diploma or less 0.444413 31% Some college but no degree 0.398918 29% 0.35808 26%
174 Table 5 6 Continue d Variable name E xp(B) e/1+e Parents 0.321422 24% Advanced degree 0.288517 22% Table 5 7. Results for multil earch stage Variable name Exp(B) e /1+e Siblings for 10 th grade student Attend college 0.926816 48% Get a job 1.023267 51% Attend trade school 1.129754 53% Join the military 1.247323 56% Get marry 1.377128 58% Whatever the student wants to do 1.52044 60% Type of college studen t plans to attend Four year college 0.926816 48% Two year college 0.546074 35% Vocational school 0.321744 24% Friends Gender Male 1.552707 61% Female 2.218874 69% Type of college student plan s to attend Four year colleg e 1.552707 61% Two year college 1.120752 53% Vocational school 0.808965 45% Relatives Gender Male 2.216657 69% Female 3.225216 76 % th grade student Attend college 2.216657 69% Get a job 2.4109 71% Attend trade school 2.622164 72% Join the military 2.851942 74% Get marry 3.101854 76% Whatever the student wants to do 3.373666 77% Type of college student plans to attend Four year college 2.216657 69% Two year college 1.488844 60% Vocational school 1 50% Do students plan to attend college, if so when Yes, after high school 2.216657 69% Yes, after one year 2.570242 72% Yes, over one year 2.980229 75% 3.455613 78% plan to continue my education 4.006828 80%
175 Table 5 7 Continue d Variable name Exp(B) e/1+e College Publications Gender Male 2.880604 74% Female 4.010837 80% Parental Education High school diploma or less 2.880604 74% Som e college but no degree 2.554867 72% 2.265963 69% 2.009729 67% Advanced degree 1.78247 64% Number of schools applied to Applied to one school 2.880604 74% Applied to two schools 3.049568 75% A pplied to three schools 3.228443 76% Applied to four schools 3.41781 77% Applied to five schools 3.618284 78% Table 5 8. Results for multil search stage Variable name Exp(B) e/1+e College R epresentatives Not fluent 7.560974 88% Partially fluent 9.225764 90% Fluent 11.25711 92% Fluency, unknown 13.73572 93% Native English speaker 16.76008 94% Type of college student plans to attend Four year college 7.560974 88% Two year college 3.807604 79% Vocational school 1.917457 66% College Source Guide Number of schools applied to Applied to one school 3.747167 79% Applied to two schools 4.026913 80% Applied to thre e schools 4.327543 81% Applied to four schools 4.650617 82% Applied to five schools 4.997811 83% None of the above sources Gender Male 0.956954 49% Female 0.703984 41% Number of schools applied to Applied to one school 0.4890 02 49% Applied to two schools 0.467297 47% Applied to three schools 0.445715 45% Applied to four schools 0.424336 42% Applied to five schools 0.403236 40% Applied to six schools 0.382488 38%
176 Table 5 8 Continued Variable name Ex p(B) e/1+e None of the above sources Type of college student plans to attend Four year college 0.956954 49% Two year college 0.398918 29% Vocational school 0.166294 14% Do students plan to attend college, if so when Yes, after hig h school 0.956954 49% Yes, after one year 1.249821 56% Yes, over one year 1.632316 62% 2.131871 68% 2.78431 74% Table 5 9. Results for multilevel logistic regression mo search stage Variable name Exp(B) e /1+e School Counselor Generational Status Immigrant 1.514371 60% First generation 1.093081 52% Second generation 0.788991 44% th grade stu dent Attend college 1.514371 60% Get a job 2.159766 68% Attend trade school 3.080217 75% Join the military 4.392946 81% Get marry 6.265134 86% Whatever the student wants to do 8.935213 90% th grade student Attend college 1.514371 60% Get a job 1.232445 55% Attend trade school 1.003005 50% Join the military 0.816278 45% Get marry 0.664314 40% Whatever the student wants to do 0.540641 35% Number of schools applied to Applied to one school 1.514371 60% Applied to two school s 1.676989 63% Applied to three school s 1.85707 65% Applied to four school s 2.056489 67% Applied to five school s 2.277322 69% Type of college student plans to attend Four year college 1.514371 60% Two year college 2.114884 68% Vocational school 2.953527 75% Teacher Number of schools applied to Applied to one school 8.828631 90% Applied to two school s 9.253483 90% Applied to three school s 9.698779 91% Appl ied to four schools 10.1655 91% Applied to five schools 10.65469 91%
177 Table 5 9. Continued Variable name Exp(B) e/1+e Parents Native Language English 0.495593 33% Spanish 0.387128 28% Other language 0.302401 23% Generat ional Status Immigrant 0.495593 33% First generation 0.640184 39% Second generation 0.826959 45% Parental Education High school diploma or less 0.495593 33% Some college but no degree 0.433008 30% 0.378326 27% 0.330549 25% Advanced degree 0.288806 22% th grade student Attend college 0.495593 33% Get a job 0.437359 30% Attend trade school 0.385968 28% Join the military 0.340616 25% Get ma rry 0.300592 23% Whatever the student wants to do 0.265272 21% th grade student Attend college 0.495593 33% Get a job 0.576373 37% Attend trade school 0.67032 40% Join the military 0.77958 44% Get marry 0.906649 48% Whatever the student wants to do 1.05443 51% Family Generational Status Immigrant 1.373003 58% First generation 1.026341 51% Second generation 0.767206 43% Table 5 10. Results for multilevel logistic regression model search stage Variable name Exp(B) e /1+e Friends Gender Male 1.999706 67% Female 2.601271 72% Teacher 0 th grade student Attend college 1.999706 67% Get a job 2.344331 70% Attend trade school 2.748348 73% Join the military 3.221993 76% Get marry 3.777264 79% th grade student Whatever the student wants to do 4.42823 82% College Sources Gender Male 7.814649 89% Female 13.91545 93%
178 Table 5 10 Continued Variable name Exp(B) e/1+e th grade student Attend college 7.814649 89% Get a job 8.550588 90% Attend trade school 9.355833 90% Join the military 10.23691 91% Get marry 11.20097 92% Whatever the student wants to do 12.25581 92% th grade student Attend college 7.814649 89% Get a job 9.914517 91% Attend trade school 12.57864 93% Join the military 15.95863 94% Get m arry 20.24687 95% Whatever the student wants to do 25.68738 96% th grade student Attend college 7.814649 89% Get a job 6.171858 86% Attend trade school 4.874415 83% Join the military 3.849718 79% Get marry 3.040433 75% Whatever the student wants to do 2.401275 71% Number of schools applied to Applied to one school 7.814649 89% Applied to two schools 9.346482 90% Applied to three schools 11.17859 92% Applied to four schools 13.36982 93 % Applied to five schools 15.99058 94% Type of college student plans to attend Four year college 7.814649 89% Two year college 12.07334 92% Vocational school 18.65287 95% Libraries No significant variables None of the above source s th grade student Attend college 1.566745 61% Get a job 1.081123 52% Attend trade school 0.746022 43% Join the military 0.514788 34% Get marry 0.355226 26% Whatever the student wants to do 0.24 5122 20% th grade student Attend college 1.566745 61% Get a job 2.272771 69% Attend trade school 3.296957 77% Join the military 4.782675 83% Get marry 6.937906 87% Whatever the student wants to do 10.06436 91% Number of schools applied to Applied to one school 1.566745 61% Applied to two schools 1.335092 57% Applied to three schools 1.13769 53% Applied to four schools 0.969476 49% Applied to five schools 0.826133 45%
179 Ta ble 5 11. Res ults for multilevel multinomial logistic regression on 12 th grade college choice stage Variable name Exp (B) e/1+e+e+e+e+e Percentage Reputation vs. Program of study Type of college Four year college 0.158025 0.056588639 6% Two year college 0.213739 0.064654802 6% Vocational school 0.289095 0.071224963 7% Cost vs. Program of study No significant variable s Location vs. Program of study Type of school Four year college 0.154278 0.05524668 6% Two year col lege 0.27999 0.084695492 8% Vocational school 0.508139 0.125191349 13% Family vs. Program of study th grade student Attend college 0.345764 0.131583079 13% Get a job 0.409835 0.143855575 14% Attend trade school 0.48578 0.15661132 16% Join the military 0.575797 0.169794141 17% Get marry 0.682495 0.183344784 18% Whatever the student wants to do 0.808965 0.197202557 20% th grade student Attend college 0.242683 0.109039101 11% Get a job 0.201897 0.098497695 10% Attend trade school 0.167965 0.088347262 9% Join the military 0.139736 0.078698153 8% Get marry 0.116251 0.069639842 7% Whatever the student wants to do 0.096714 0.061237915 6% Other reason vs. P rogram of study th grade student Attend college 0.426135 0.179944211 18% Get a job 0.358438 0.154774694 15% Attend trade school 0.301496 0.132513733 13% Join the military 0.253599 0.11298833 11% Get marry 0 .213312 0.095988003 10% Whatever the student wants to do 0.179425 0.081281837 8% Table 5 12. Results for bi college enrollment Variables name Exp (B) e/1+e Number of schools applied to Applied to o ne school 0.373439 27% Applied to two school s 0.569498 36% Applied to three school s 0.868489 46% Applied to four school s 1.324454 57% Applied to five school s 2.019803 67% Type of college student plans to attend Four year college 0. 373439 27% Two year college 0.537944 35% Vocational school 0.774916 44% High school type Public school 0.373439 27% Catholic school 0.183232 15% Private school 0.089905 8%
180 Table 5 13. Summary of significant independent variables for each multivariate logistic regression model Independent Variables Model 1 Predispositio n Stage Model 2 Search Stage Model 3 Search Stage Model 4 Choice Stage Model 5 College Enrollment Demographic Characteristics Gender X X X Family composi tion X Number of in home siblings Social & Cultural Capital X X X X X obtainment X X Social n etworks expectation fo r 1 0 th grader after high school X X X Social n etworks expectation for 12 th grader after high school X X Supply of Resources Family income Number of schools applied to X X X X Expected Costs Type of c ollege X X X X X When stude nts plan to enroll in college X X High school Context High school type X Degree of urbanization
181 C HAPTER 6 IMPLICATIONS AND FUTURE RESEARCH C hapter 6 will introduce key institutional recommendations related to Latina/o studen S uggestions for future research on this context area will be explored to assist in the continue d examination of Latina/o and first generation Finally, C hapter 6 will close with final thoughts re lated to thi s study and the challenges students must overcome in order to successful ly n avigat e the college choice process. Gaining access to postsecondary education is often seen as a gateway to a variety of other benefits, such as higher sal aries healt hier lifestyles and civic involvement (Baum & Ma, 2007). I n comparison to high school students, college graduates earned over twenty thousand dollars more i n annual salary (Baum, et al., 2010). R ecent evidence s suggest that adults who earned at most a hig h school diploma were three times more likely to smoke than adults who earned a b (NCHS, 2012). These findings highlight the advantages that emerge when people continue their education beyond the high school level and complete a college deg ree This f urther e mphasiz es the importance of assisting Latina/o s especially first generation Latina/o students in pursuing a college degree Although first generation Latina/o students are increasing their numbers in college enrollment (Choy, 2001; Str ayhorn, 2006), they still encounter various struggles in college access and retention In particular, first generation students continue to demonstrate lower degree aspirations than many of their non first generation peers (Lohfink & Paulse n, 2005; Saenz, et al., 2007). This may suggest that first generation students are questioning their ability to earn a college degree This study highlight s characteristics and other fa ctors that
182 positively influence ollege choice process S ubsequently it pro vide s con text for institutional recommendations to help o enroll in college and earn a college degree. Implications for Policy and Institutional Practice s This study generated numerous suggestions for policy implic ations relevant to the development of In particular, one of the most significant findings highlight that first generation students often seek college information from their parents. First generation students may b enefit more in receiving accurate information about college since many of thei r parents do not have first hand ex perience of the college process To better facilitate first generation students and rative for a college going s Previous studies have found that students who attend schools where going to four year institutions is the norm are more likely to pursue college opportunities (Engberg & Wolniak, 2009, 2010; Hossler, et al., 1999; McDonough, 1997). Therefore, two institutional practices will be recommended for this study that support the idea of creating a college going culture at high schools and the Latino community The i ntention of t hese recommendations is to educat e Latina/o families about the college choice process therefore they can make informed decisions about college. State and district policy requiring schools to develop college preparation programs for high schools. State an d district policy requiring community outreach and partner ship with Spanish language media outl ets to share with the Hispanic community information on the pathway to college
183 Institutional Practice s and Recommendations These programs should start when st udents first enter high school since (Hossler, et al., 1999). As previously mention ed it is imperative for first generation Latina/o students to gather college information early on to successfully navigate the college choice process and develop informed educational aspirations. While it is not possible to provide a complete step by step plan for the institutional practices, high school administrators can use the following r ecommendations to develop programs that will assist first generation students to better understand the college choice process. The first institutional practice is for school districts to deve lop a college preparation program particularly, at s chools with a large Latina/o population The college preparation programs should offer three components: college prep courses, colleg iate opportunities and informative college sessions for parents. A college going culture supports behaviors that relate to students pre paring for and applying to college. While college preparation programs are often found in specialized areas of high schools, such as magnet or advance placement programs, they should ideally be accessible to all students (Corwin & Tierney, 2007), not only gifted students. Each school district should determine what is taught in the elective courses however, schools should have the flexibility to alter the curriculum ba sed on their student population (i.e., first generation, migrant, low income). There shou ld be two elective courses off ered, one geared towards lower classman and another for upper classman The purpose fo r this structure is to eliminate redundancy between both courses and cover more appropriate topics geared towards each grade level. For exam ple, the first college prep program should introduc e basic college knowledge, such
184 as the importance of a college degree and the requirements needed to be admitted to college. I mportant topics to be discussed as support by th e findings in this study, shou ld include the various type s of postsecondary institutions available the importance of applying to multiple colleges, and the different ways to finance a college education. In particular, conveying to students eligibility to state wide scholarships, su ch Bright Futures s st Century Scholars p rogram, so the y can prepare and ensure t o meet the requirements by the ir senior year In this course, students will also be informed of the many r esources available to them on the local, state, and national level The second college prep course should incorporate more detail information pertaining to college deadlines requirements for the college a pplication cycle, and how to apply for v arious financial aid support This course s hould be more interactive and grant s tudents the opportunity to work on actual college and financial aid applications. As well as, completing scholarship applications and learning the various styles towards writing a college or scholarship essay. Since m an y first generation students lack information pertaining to the cost of postsecondary institutions (Hsiao, 1992; L. Perna, 2004; York Anderson & Bowman, 1991) more advanced workshops sho uld train school counselors and teachers about the nuances of the finan cial aid process and the many financial opportunities As supported by t he findings in this study teachers and school counselors were the most sought out for college information (Horn & Nunez, 2000; Saenz, et al., 2007). Therefore, school counselors shoul d be more knowledgeable of the admissions requirements of various types of postsecondary institutions to be better equipped to advise diverse student groups about potential college options.
185 Secondly, s c hool s should provide collegiate opportunities for all students High school s eniors should be require to attend a minimum of c ollege readiness acti vities P revious college enrollment programs mandated by the government have mainly focused on providing financial support to students H owever, th is college prepa ration program should mainly focus in preparing students for their future educational endeavors (L. Perna, Rowan Kenyon, Bell, Thomas, & Li, 2008). These c ollegiate opportunities should include c ollege tours college fairs financial aid workshops writing sessions, and college su m mits Ideally, h igh schools should form partnerships with a variety of postsecondary institutions (e.g., selective, liberal arts, Hispanic serving institutions, state colleges, community colleges, research intensive universities e tc.) and have college representatives such as, admission recruiters, financial aid advisors, a nd a cademic advisors speak to students a bout their institution This component is vital to this study because it found that college sources, such as college publi cations, guides, and representatives, were the most sought out by Latina/o students for college informatio n. Through these collegiate opportunities, students will have direct connect with various college sources to help enhance their college choice process Exposing students to a variety of postsecondary institutions may encourage many who originally planned to enroll at a two year college to reconsider a four year institution especially once they hear the benefits of attending a four year college In ad dition, this program will provided students the opportunit y to interact with others and converse about college related topics. These interactions should be with fellow classmates, teachers, and counselors. According to the findings, friends, teache rs, and counselors were found to be the most significant
186 in their college search stage to attend college. Moreover, school counselors and teachers should be reminded to be mindful of the covert and overt messages they send to students about their college readiness and abilities. Particularly, toward s low income, first generation or Latina/o s tudents who often need additional support and encouragement from institutional gatekeepers compared to students that come from more affluent backgrounds Based on th e se findings, school counselors, coaches, and teachers may need to be more intentional, inf ormative, and direct in their messages to students about college expectations. Teachers and counselors can begin to develop and express high expectations for Latina/o s tudents through school related activities that promote a college going culture. O verall, th is college knowledge program must work to create a school and counsel ing environment that nurture s generation Latina/o students (McClaffery, McDonough, & Nunez, 2002). According to this study gender was another significant variable found in Latina/o higher degree aspirations and seek various sources for college information. For this reason, a college su m mit that is specifically for Latina/o men or men of color should be created as an additional source of support and encouragement for this population. Whether individual schools or the district as a whole should sponsor such event should be left for the distri ct to decide. Overall, the college su m mit for males should have inspirational male speakers and role models that male students can relate to and would
187 want to aspire to be like. As well as fun and creative college readiness activities and sessions that cap In regards to Latina/o parents, it is imperative they receive accurate knowledge about college in a timely manner. The college knowledge programs must be develop to fit the unique needs of Latina/o parents in order to success fully disseminate college information to them. For example s chools should host meetings available to parents at convenient location s and with the presence of a Spanish translator. By providing bilingual meetings that are not restricted to a certain time o r day, more parents may attend and pass down the information they acquired from these sessions to their children or o ther Latina/o families. Additionally, s chool counselors must be willing to alter the information sessions to meet the needs of certain pare ntal populations (e.g., migrant families, recent immigrants, undocumented households etc. ). I nforming parents about the various opportunities offered to their children such as advanced placement courses, dual enrollment, and other college geared programs despite some legal status may encourage more Latina/o parents to push their children to take advantage of these opportunities and earn a college degree Schools should also e mphasiz e to Latina/o parents about the importance of talking to and e ncouraging t heir children to attend college. These dialogue between parent and student will help to make a difference of how students navigate the college choice process. As found in this study, first generation Latina/o students were more likely to seek c ollege infor mation from their parents than students who had college educated parents. Therefore, school administrators should track those parents that attend the information sessions and follow up with th os e who have not attend ed This
188 will help to e nsure that first generation families are obtaining the necessary information to help their children get into college P revious research has noted that first generation students often lack information related to the costs of postsecondary institutions (Hsiao, 199 2; L. Perna, 2004; York Ande rson & Bowman, 1991). To improve the likelihood that first generation students and families g ather the co rrect information about financial obligations pertaining to p ostsecondary institutions, school districts should mandate t ha t all ninth grade students and parents receive informational packet s on ways to prepare, save, and pay for college through the college preparation programs These informational packets should be disturbed annually throughout student s high school career to keep families informed about the importance of preparing for college. Information about the various types of financial aid options available to students, along with the limitations and alternatives (e.g., D REAM Act) is important to include This informat ion can be disbursed to parents via mail or at informational sessions either on or off campus In accordance to federal regulations, information sent home should be translated to the language spoken in the student s household. Having a Spanish speaking tra nslator at these m eetings will help to encourage that more non English speaking parents attend such session The information provided to parents and students should focus on the big picture and not the complex details that may overwhelm most individuals F or example, a two page document should list the requirements of the different type of institutions and the average cost for community college s versus public and private four year college that are either located nearby or throughout the state. The cost shou ld highlight what the listed tuition is a long with any additional costs A s eparate column should display what the average student pays after
189 federal g rants and financial aid are taken into consideration. In addition, a clear overview of the various finan cial aid options available to students such as scholarships, grants, loans, and work study should be provided I nformation for L atina/o students who find themselves in complicated sit uations should be discussed in particular students whose parents are u ndocumented For example, in Florida many colleges charge American born students whose parents are undocumented out of state fees, despite the residency in this status Thus, Latina/o students who find themselves in similar situatio ns should be informed of alternative methods such as, schools that offer in state tuition and funding The policy and institutional recommendations highlight key areas that must be developed in order for schools to provide students with assistance especi ally first generation students who have a variety of disadvantages placed against them. Wh ile it may be ideal for all these recommendations to be implemented on a widespread basis, I recognize cost and time may prohibit the implementations of all of them To address limited fund ing school districts and schools should assess which recommendations would be most meaningful at their school and implement th e ones that will have the largest impact. Nonetheless, schools that strive to develop a college going cult ure should provide an environment that assist all students in recognizing that college is an opportunity available to students of all races and backgrounds. Partnership between Spanish Language Organizations The final policy recommendation is related to community outreach and partnership between school districts and Spanish language media outlets regarding the dissemination of college informati on with in the Hispanic community E xtant literature has continuously highlighted the significant role p lay in the development of
190 s avigat ion in t he earliest stages of the college choice process (Chapman, 1981; Flint, 1992; Hossler, et al., 1989) Therefore, parents should ha ve access to college information, such as college preparation programs, funding college and accessibility to local resources This type of information should be televised and broadcast ed regularly on Spanish stations in order to inform the Hispanic communities about college requirements, deadlin es, a nd resources available to students T he benefits of mandating such partnerships is to inform a larger Hispanic audience that may prove invaluable to Latina/o students especially those who otherwi se may not gather college infor mation from their parents Additionally, community outreach to Hispanic institutes, such as religious institutions or Hispanic establishments, will help to spread the word to the Latina/o community about the importance of a college education. I nformation provided to the larger Hispanic community should consist of the necessary steps needed to enroll in college wa ys to finance a college education and opportunities and resources available to high s chool s tudents. The collaboration between Hispanic institut ions and secondary schools will entail the assistance of a variety of constituents that will help to create a college going culture within the Latina/o community Schools should participate in Hispanic events or functions as a way to engage with the Hispan ic c ommunity and make their presence visible. Future Research There are many d ifferences that exist within the Latina/o student population as it r elate s to college plans (Olive, 2008; Phinn ey, et al., 2005) unfortunately, limited information acknowledges t his group as a heterogeneous population. A dearth of evidence exists on the preparation methods Latina/o students utilize when trying to gain
191 access to postsecondary institutions (Gofen, 2009). While this study provides insight on the within group differe nces of Latina/o students, further research is necessary to better college choice process and the pathways they take to gain access t o postsecondary institutions. B here are five mai n areas that need furt her research to strengthen Latina/o college choice proces s : the role of social network s, postsecondary institutions, college applications, gender, a nd m ethodological modifications. Social Networks 006) proposed conceptual model choice process. While this relatively new theoretical framework has been tested on limited research (Bergerson, 2009), it was relevant for social networks because it incorporates ideas of social capital (e.g., information about college). While this study did choice process, future studies that use a di fferent dataset need to expanded on the number of social capital variables included. The first variable that should be included in future studies is the actual expectations social networks have for student s what t hey think those individuals desire for them to do after high school. More specifically, a mix method study would be more helpful in examining the expectations school counselors, teachers, friends, and parents have for Latina/o students and the reasons why. This information will provide d further explanation for the expectations social network s have for that particular student. Although it may be more time consuming and expensive, this type of study may be more informative and helpful for
192 quantitative data. A dditional research (i.e., qualitative and quantitative) should also examine the level of influence social networks have on the college choice process of students of various backgrounds in order to observe any similarities or differences across different ra cial groups. networks to facilitate educational advancement for students (Bourdieu, 1986; S tanto Salazar, 2004), a clearer definition for such individuals should be clarified. For example, college choice process. However, many may confuse the term social netwo rk with social media via internet Whereas terms such as, institutional gatekeepers and educational agents, are not inclusive of parents or friends. Therefore, future studies should consider using a more inclusive, yet specific term, that pertains to the e xchange of information, resources, and support from social networks towards student college choice process Gender, College Type and C ollege Applications The additional variables that were found the most significant towards Latina/o hoice process was gender, the type of postsecondary institution students planned to enroll in, and the number of college applications they submitted. While there is an abundance of research that has difference s in referen ce to college aspirations college enrollment and college graduation rates (Zarate & Gallimore, 2005; King, 2006), more studies should e xclusively focus on choice process (Saenz & Ponjuan, 2 009). Secondly, the type of institutions students
193 aspire d choice process and enrollment Especially since this was the only variable to be significant across all five models and due to Latina/o students high aspirations to attend either a two year college or vocational school This is problematic because s e veral studies have at a community college decreases their gree (Admon, 2006; Laden, 2004). However, limited research has examined Latina/o Thus, r esearching when and why Latina/o students are more likely to aspire to attend a v ocational or two year college will p rovide d helpful insight of what is impacting these aspirations. Especially, since o ver half (59%) of the Latina/o students in this sampe planned to attend a vocational technical trade school versus only 5% who planned to attend a four year college/universi ty. found significant across all but one model. This variable should be further study, exclusively or in correlation to other related variables, in order to better understand how college income, college awareness, academic preparation, and college readiness. Methodological Modifications The final area to be discussed will address some methodologic al modificati ons that could strengthen this study and provide more robust insights regarding the college choice process of Latina/o students. Every study inherently has limitations as a result of the data being used and the knowledge of the researcher, but exploring ways to minimize these limitations will improve results and provide more accurate assessments o f the problem under study For future studies on the college choice process of Latina/o
194 students, there are t wo methodological modificati ons that shou ld be considered: survey used and type of data. First, this study used the EL S:0 2 dataset and was subsequently limited to the que stions included in the survey. While the E LS:0 2 dataset includes surveys administered to students, parents, administrators, and school librari an s the questions included in the survey were not developed to study college choice process. Future studies should consider using an alternative survey or developing one that is focused on the nuances that have emerged from previous studies on first generation and Latina/o students. The E LS:0 2 dataset was an effective survey for this study but a more robust study that focused on college choice process would be more suitable and m ay provide more insights into this topic Also, this dataset was a ccessible to the public. H ad I requested a license version of thi s dataset additional variables and i nformation would have been provi d ed and manipulated and controlled for Thus, the unlicen sed dataset wa s prone to biasness from the statisticians who created the public version of ELS:02. As previously noted, using quantitative data is the norm when examining college choice and college enrollment (L. Perna, 2006), but by only incorporating qua ntitative data the results may exclude the other half of the story T here is a dearth of evidence on the college choice process of first generation Latina/o students (Gofen, 2009; Thayer, 2000), particularly on why some Latina/o s are successful in navigati ng the college choi ce process while others are not. Further research should focus on the reasons why some Latina/o students choose to pursue a college edu cation, along with the methods they use to successfully navigate the college choice process To examin e these areas,
195 qualitative research may be best suited to gain this background information and the re sults can be later translated into a more appropriate quantitative methodology. C losing Words Chapter 6 further explore s specific findings from this stud y in collaboration to future recommendations and research Specifically, the role of demographic characteristics, cultural and social capital, supply of resources, expected cost, and school contex t were examined to determine their role on Latina/o students college choice process and enrollment With a pproximately one third of all children who are enrolled in s econdary school com e from ho useholds where the parents first hand experience of the college choice process (Aud et al., 2012), President O bama is urging more individuals to earn a postsecondary degree F or this to be made possible, additional policies and programs must be put in place with in secondary schools T herefore, I recommended a state and district policy to mandate high schools to d evelop a college preparation program so more Latina/o students can learn about the steps needed to enroll in postsecondary institutions. This college preparation program should incorporate informational college prep cour ses, collegiate opportunities, colle ge summits, and informative college sessions for parents. Additionally, I recommended a state and district policy to require community outreach and partnership between high schools and Spanish language media outlets to share college information to the Hisp anic community. Overall, these two recommendations will help to create a college going culture at both the high school s students attend and the communities they live in Researchers should continue to fill the current gaps in the literature regarding Latin a/o developing these critical policies and programs. Future research studies should
196 examine the role of social networks, different type of postsecondary institutions, num ber of college applications and gender in relations to La college choice process since these four variables were the most significant in this study.
197 REFERENCES Adelman, C. (2009). The spaces between numbers: Gettin g international data on h igher education straight. Washington, DC. Adelman, C., Daniel, B. & Berkovits, I. (2003). Postse condary attainment, attendance, curriculum, and performance: Selecte d results from the NELS:88/2000 postsecondary education transcript study (PETS), 2000, (NC ES Publication No. 2003 394). Washington, DC: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics. Admon, N. (2006). Hispanic students and the decision to attend a community college (Doctoral dissertatio n). Available from ProQuest Dissertations & Theses database. (AAT 3235692) Alon, S., & Tienda, M. (2005). Assessing the "mismatch" hypothesis: Differences in college graduation rates by institutional selectivity. Sociology of Education, 78 (4), 294 315 Am erican Community Survey (2008). Selected population profile in the United States Hispanic or Latino (1 year estimates) Washington, DC: U. S. Census Bureau. American Institutes for Research (2010). Finishing the first lap: The cost of first year student year colleges and universities Washington, DC: AIR. American School Counselor Association. (n.d.). Student to counselor ratio by state: 2009 10 Retrieved from http://www.schoolcounselor.org/files/Ratios09 10.pdf Anderson, V. (2008). The influence of race based rejection sensitivity on the college choice and freshman year transition of Black and Hispanic high school students (Doctoral dissertation). Available from ProQuest Dissertations & Theses database. (AAT 3333296) Arbona C., & Nora, A. (2007). The influence of academic and environmental factors on Hispanic college degree attainment. The Review of Higher Education, 30 (3), 247 269. doi: 10.1353/rhe.2007.0001 Astin, A.W. (1973). The impact of dormitory living on students. Educational Record, 54, 204 210. Astin, A.W. (1999). Student involvement: A developmental theory for higher education. Journal of College Student Development, 40 (5), 518 529. Aud, S., Hussar, W., Johnson, F., Kena, G., Roth, E., Manning, E., & Zhang, J. (2012). The Condition of Education 2012 (NCES 2012 045). Washington, D.C.: U.S Department of Education, National Center for Education Statistics.
198 Azmitia, M., & Cooper, C. R. (2001). Good or bad? Peer influences on Latino and European American adolescen ts' pathways through school. Journal of Education for Students Placed at Risk, 6 (1/2), 45 71. Azmitia, M. et al. (1994). Links between home and school among low income Mexican American and European American families UC Berkeley: Center for Research on Education, Diversity and Excellence. Retrieved from: h ttp://www.escholarship.org/uc/item/1n9843b7 Badger, E. (2010, May). Minding the education gap. Miller McCune.com. Retrieved from http://www.miller mccune.com/business economics/minding the educationg ap 16074/# Baum, S., & Ma, J. (2007). Education pays: The benefits of higher education for individuals and society. College Board. Access from http: //www.collegeboard.com/prod_downloads/about/news_info/cbsenior/yr2007/ ed pays 2007.pdf Baum, S., Ma, J., & Payea, K. (2010). Education pays 2010: The benefits of higher education for individuals and society Washington, D.C.: The College Board. Behnke, A. O., Piercy, K. W., & Diversi, M. (2004). Educational and occupational aspirations of Latino youth and their parents. Hispanic Jounral of Behavioral Sciences, 26 16 35. Bell, A., Rowan Kenyon, H., & Perna, L. (2009). College knowledge of 9th and 11th grade students: Variation by school and state context. The Journal of Higher Education, 80 (6), 663 685. Bentez, M., & DeAro, J. (2004). Realizing student success at Hispanic serving institutions. New Directions for Community Colleges, 127 35 48. doi:1 0.1002/cc.162 Bergerson, A. A. (2009a). Introduction to college choice. ASHE Higher Education Report, 35 (4), 1 10. doi: 10.1002/aehe.3504 Bergerson, A. A. (2009b). College choice as a comprehensive process. ASHE Higher Education Report, 35 (4), 21 46. Be rkner, L., et al. (2005). National Postsecondary Student Aid Study (NPSAS:04), 2003 04: financial aid estimates for 2003 04 by type of institution. E.D. TAB, National Center for Education Statistics, U.S. Department of Education NCES 2005 158, Table 1. Bers, T. (2005). Parents of traditionally aged community college students: Communications and choice. Research in Higher Education, 46 (4), 413 436. doi: 10.1007/s11162 005 2968 z
199 Bhagat, G. S. (2004). The relationship between factors th at in fluence college choice and persistence in Longhorn Opportunity Scholarship recipients at the University of Texas at Austin (Doctoral dissertation). Available from ProQuest Dissertations & Theses database. (AAT 3150544) Bourdieu, P. (1986). The forms of ca pital. In J.G. Richardson (Ed.), Handbook of theory and research for the sociology of education (pp.241 258). New York: G reenwood. Bourdieu, P., & Passeron, J. C. (1990). Reproduction in education, society, and culture (1990 ed.). London; Newbury Park, CA : Sage Publications. economic resources, employment, and living arrangements. Sociology of Education, 80 (3), 261 284. Brick, K., Challinor, A.E., & Rosenblum, M.R. (20 11) Mexican and Central American immigrants in the United States. Migration Policy Institute. Accessed from http://www.migrationpolicy.org/pubs/mexcentamimmigrants.pdf Brown, H. E., & Burkhardt, R. L. (1999, May). Predicting student success: The relativ e impact of ethnicity, income, and parental education. Paper presented at the Annual Forum of the Association of Institutional Research, Seattle, WA. Brown, S.E., Santiago, D. & Lopez, E. (2003). Latinos in higher education: Today and tomorrow. Change: T he Magazine of Higher Learning. Bryan, J., Moore Thomas, C., Day Vines, N., & Holcomb McCoy, C. (2011). School counselors as social capital: The effects of high school college counseling on college application rates. Journal of Counseling and Development 89, 190 199. Bui, K. (2002). First generation college students at a four year university: Background characteristics, reasons for pursuing higher education, and first year experiences. College Student Journal, 36 (1), 3 11. Burdman, P. (2005). The student debt dilemma: Debt aversion as a barrier to college access. UC Berkeley: Center for Studies in Higher Education Retrieved from: http://escholarship.org/uc/item/6sp9787j Butner, B., Caldera, Y., Herrera, P., Kennedy, F., Frame, M., & Ch ilders, C. (2001). The college choice process of African American and Hispanic women: Implications for college tran sitions. The Journal of College Orientation and Transition, 9 (1), 24 32. Cabrera, A., & La Nasa, S. (2000a). Overcoming the tasks on the path to college for America's disadvantaged. New Directions for Institutional Research, 107 31 43.
200 Cabrera, A., & La Nasa, S. (2000b). Understanding the college choice process. New Directions for Institutional Research, 107 5 22. Cabrera, A. F., & La Nasa, S. M. (2001). On the path to college: Three critical tasks facing America's disadvantaged. Research in Higher Edu cation, 42 (2), 119 149. Carreras, I. E. (1998). Institutional characteristics of importance at the college search stage among Latino high school students (Doctoral dissertation). Available from ProQuest Dissertations & Theses database. (AAT 9835470) Cart er, D. F. (2003). College students' degree aspirations: A theoretical model and literature review with a focus on African American and Latino students. In J. C. Smart & W. G. Tierney (Eds.), Higher education: Handbook of theory and research (Vol. XVII, pp. 129 171). Netherlands: Kluwer Academic Publishers. Castillo, L.G., Conoley, C.W., Archuleta, D.J., Phoummarath, M.J., Landingham, A.V., Choi Pearson, C. (2006). University environment as a mediator of Latino ethnic identity and persistence attitudes. Jou rnal of Counseling Psychology 53 (2), 267 271. DOI: 10.1037/0022 0188.8.131.527 Ceja, M. (2001). Applying, choosing, and enrolling in higher education: Understanding the college choice process of first generation Chicana students. (Doctoral dissertation), University of California, Los Angeles. Dissertations & Theses. (AAT No. 3026250) Ceja, M. (2006). Understanding the role of parents and siblings as information sources in the college choice process of Chicana students. Journal of College Student Developm ent, 47 (1), 87 104. Cellini, S.R. (2008). Causal inference and omitted variable bias in financial aid research: Assessing solutions. The Review of Higher Education, 31(3), 329 354. Census Bureau. (2011). Educational attainment by race and Hispanic orig in: 1970 to 2010. U.S. Census Bureau, Statistical Abstract of the United States. Retrieved May 12, 2013. http://www.census.gov/compendia/statab/2012/tables/12s0229.pdf Census Bureau, (2012). Income, Poverty, and Health Insurance Coverage in the United States: 2011. Retrieved May 12, 2013 http://www.census.gov/prod/2012pubs/p60 243.pdf Chapman, D. W. (1981). A mod el of student college choice. Journal of Higher Education, 52 (5), 16. Chen, X. (2005). First generation students in postsecon dary education: A look at their college transcripts. (NCES 2005 171). Washington, D.C.: U.S. Government Printing Office.
201 Choy, S (2001). Students whose parents did not go to college: Postsecondary access, persistence, and attainment. U.S. Department of Edu cation, Institute of Education Sciences, National Center for Education Statistics. Cohen, D. (2009). Relationships that create confidence: Understanding postsecondary academic choices of Mexican heritage high school graduates in light of influential relationships, self efficacy, and mathematical experiences (Doctoral dissertation). Available from ProQuest Dissertations & Theses d atabase. (AAT 3355733) Coleman, J.S. (1988). Social capital in the creation of human capital. The American Journal of Sociology, 94, S95 S120. Collatos, A., Morrell, E., Alejandro, N., & Lara, R. (2004). Critical sociology in K 16 early intervention: Re making Latino pathways to higher education. Journal of Hispanic Higher Education 3 (2), 164 179. Contreras Godfrey, R. (2009). Giving voice to Black and Latino men: First year students' perceptions of the relative impact of family support and college a spirations on their decisions to enroll and actual college enrollment (Doctoral dissertation). Available from ProQuest Dissertations & Theses database. (AAT 3349653) Corwin, Z. B., & Tierney, W. (2007). Getting There -And Beyond: Building a Culture of Col lege Going in High Schools. Center for Higher Education Policy Analysis, University of Southern California 32. Cruz V., Capps, R., Vericker, T., & Kuehn, D. (2009) Educational attainment of f irst and s econd g eneration i mmigrant y outh. Urban Institute Davis, J. (2010). The first generation student experience: Implications for campus practice, and strategies for improving persistence and success Sterling, VA: Stylus Publishing, LLC. DiMaggio, P., & Mohr, J. (1985). Cultural capital, educational atta inment and marital selection. The American Journal of Sociology, 90 (6), 1231 1261. 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. Durand, J., & Massey, D.S. (1992). Mexican migration to the United States: A critical review. Latin American Research Review, 27 (2), 3 42. Engberg, M. E., & Wolniak, G. C. (2009). Navigating disparate pathways to college: Examining the conditi onal effects of race on enrollment decisions. The Teachers College Record, 111 (9), 2255 2279.
202 Engberg, M. E., & Wolniak, G. C. (2010). Examining the effects of high school contexts on postsecondary enrollment. Research in Higher Education, 51 (2), 132 15 3. doi: 10.1007/s11162 009 9150 y Engle, J., & Tinto, V. (2008). Moving beyond access: College success for low income, first generation students Washington, DC: The Pell Institute for the Study of Opportunity in Higher Education. Ennis, S.R., Rios Varga s, M., & Albert, N.G. (2011). The Hispanic population: 2010. U.S. Census Bureau, 2010 Census Briefs. Retrieved May 1, 2012. http://www.census.gov/prod/cen2010/briefs/c2010br 04.pdf Federal Student Aid. (2010). Dependency status worksheet. Retrieved from http://www.fafsa.ed.gov/FOTWWebApp/fotw1011/WorksheetServlet?locale=en_ US&wstyp e=WSDEP Feliciano, C. (2005). Does selective migration matter? Explaining ethnic disparities in educational attainment among immigrant's children. International Migration Review, 39 (4), 841 871. doi: 10.1111/j.1747 7379.2005.tb00291.x Fitzgerald, B. (2004). Federal financial aid and college access. In E. P. St. John (Ed.), Public policy and college access: Investigating the federal and state roles in equalizing postsecondary opportunity (Vol. 19, pp. 1 28). New York, NY: AMS Press, Inc. Fry, R. (20 02). Latinos in higher education: Many enroll, too few graduate. Pew Hispanic Center, 22 (3), 1 32. Fry, R. & ERIC Clearinghouse on Urban Education, N. Y. (2003). Hispanics in College: Participation and Degree Attainment. Retrieved from ERIC database. ( ED480917) Fry, R. (2007). Are immigrant youth faring better in U.S. schools? The International Migration Review, 41 (3), 579 601. Fry, R. (2008). Latino settlement in the new century Washington, DC: Pew Hispanic Center. Fry, R. (2012). Hispanic colleg e enrollment spikes, narrowing gaps with other groups. Washington, DC: Pew Hispanic Center. Gandara, P. (1995). Over the Ivy Walls: The Educational Mobility of Low Income Chicanos Albany, N.Y.: State University of New York Press. Ganderton, P. T., & Santos, R. (1995). Hispanic college attendance and completion: Evidence from the high school and beyond surveys. Economics of Education Review, 14 (1), 35 46. Doi : 10.1016/0272 7757(94)00034 4
203 Ginorio, A., & Huston, M. (2001). Â¡ Si, se puede! Yes, we can. L atinas in school Washington, DC: AAUW Foundation Glick, J. E., & White, M. J. (2004). Post secondary school participation of immigrant and native youth: The role of familial resources and educational expectations. Social Science Research, 33 (2), 272 299. doi: DOI: 10.1016/j.ssresearch.2003.06.001 Gloria, A. M. (1997). Chicana academic persistence: Creating a university based community. Education and Urban Society, 30, 107 121. Gloria, A. M., & Rodriguez, E. R. (2000). Counseling Latino university studen ts: Psychosociocultural issues for consideration. Journal of Counseling and Development, 78 (2), 145 154. Gloria, A.M, & Segura Herrera, T.A. (2004). Ambrocia and Omar go to college: A psychosociocultural examination of Chicana/os in higher education. I n R.J. Valazquez, L.M. Arellano, & B.W. McNeil (Eds.), The handbook of Chicana/o psychology and mental health (pp.401 425). Mahwah, NJ; Lawrence Erlbaum. Gofen, A. (2009). Family capital: How first generation higher education students break the intergener ational cycle. Family Relations, 58 (1), 104 120. Gomez, G. G. (2005). The negotiation to college: Examining the school and home influences in the college choice process for Mien American female and male students (Doctoral dissertation). Available from Pro Quest Dissertations & Theses database. (AAT 3190471) Gonzalez, J. (2010). Low Hispanic graduation rates threaten Obama's college attainment goal. The Chronicle of Higher Education Retrieved from http://chr onicle.com/article/Low Hispanic Graduation Rates/64710/. Gonzlez, K. P., Stoner, C., & Jovel, J., E. (2003). Examining the role of social capital in access to college for Latinas: Toward a college opportunity framework. Journal of Hispanic Higher Educat ion, 2 (2), 146 170. Gonzalez, L. (2007). College choice of Latino high school students: Influence of demographics, academic preparation, and academic self efficacy beliefs on intended level of post secondary institution (Doctoral dissertation). Available from ProQuest Dissertations & Theses database. (AAT 3306587) Graham Clay, S. (2005). Communicating with parents: Strategies for teachers. School Community Journal, 15 (1), 117. Griffith, J. (2000). School climate as group evaluation and group consensus: Student and parent perceptions of the elementary school environment. T he Elementary School Journal 101(1), 35 61.
204 Hagy, A. P., & Staniec, J. F. O. (2002). Immigrant status, race, and institutional choice in higher education. Economics of Education Revie w, 21 (4), 381. Hahs Vaughn, D. L. (2004). The impact of parents' education level on college students. Journal of College Student Development, 45 (5), 483 500. Hahs Vaughn, D. L. (2007). Using NCES national datasets for evaluation of postsecondary issues. Assessment & Evaluation in Higher Education, 32 (3), 239 254. Hahs Vaughn, D. L., McWayne, C. M., Bulotsky Shearer, R. J., Wen, X., & Faria, A. M. (2011). Methodological considerations in using complex survey data. Evaluation Review, 35 (3), 269 303. Hak imzadeh, S., & Cohn, D. (2007). English usage among Hispanics in the United States. Pew Research Hispanic Center. Hamrick, F., A., & Stage, F., K. (2004). College predisposition at high minority enrollment, low income schools. Review of Higher Education, 27 (2), 151 168. Hanson, S. (1994). Lost talent: Unrealized educational aspirations and expectations among US youths. Sociology of Education, 67 (3), 159 183. Hao, L., & Bonstead Bruns, M. (1998). Parent child differences in educational expectations and t he academic achievement of immigrant and native students. Sociology of Education, 71 (3), 175 198. Harwell, M., & LeBeau, B. (2010). Student eligibility for a free lunch as an SES measure in education research. Educational Researcher, 39 (2), 120 131. Hear n, J.C. (1988). Attendance at higher cost colleges: Ascribed, socioeconomic, and academic influences on student enrollment patterns. Economics of Education Review,7 (1), 65 76. Hearn, J. C. (1991). Academic and nonacademic influences on the college des tinations of 1980 high school graduates. Sociology of Education, 64 (3), 158 171. Hearn, J. C. (1992). Emerging variations in postsecondary attendance patterns: An investigation of part time, delayed, and nondegree enrollment. Research in Higher Education, 33 (6), 657 687. Herzog, M. J. R., & Pittman, R. B. (1995). Home, family, and community: Ingredients in the rural education equation. Phi Delta Kappan, 77 113 119.
205 Hickson, A. (2012, October 9). Florida ruling for DREAMers and children of undocument ed to receive in state tuition. NBCLatino Retrieved from http://nbclatino.com/2012/10/09/florida ruling for dreamers and children of undocumented to receive in state tuition/ Hispanic Alliance Incorporated. (2010). Fact sheet on Hispanic education Retrieved from http://ochla.ohio.gov/ASSETS/907EB369F108449EB97D1D7E22724EF9/Hispani c Alliance Fact Sheet Education FINAL.pdf Hoover Dempsey, K., & Sander, H. (1995). Parental involvement in children's education: Why does it make a difference? The T eachers College Record, 97 (2), 310 331. Horn, L., Cataldi, E. F., & Sikora, A. (2006). Waiting to attend college: Undergraduates who delay their postsecondary enrollment (NCES Publication No. 2005152). Washington, DC: U.S. Department of Education, Instit ute of Education Sciences, National Center for Education Statistics. Retrieved from: http://nces.ed.gov/pubs2005/2005152.pdf Horn, L., & Nunez, A. M. (2000). Mapping the road to college first genera tion students' math track, planning strategies, and context of support (Statistical Analysis Report No. NCES 2000 154). Washington, D.C.: U.S. Department of Education, National Center for Education Statistics Hossler, D., Braxton, J., & Coopersmith, G. (1989). Understanding student college choice. In J. C. Smart (Ed.), Higher education: Handbook of theory and research (Vol. IV, pp. 231 288). New York, NY: Agathon Press. Hossler, D., & Gallagher, K. (1987). Studying college choice: A three phase model an d the implications for policy makers. College and University, 2 207 221. Hossler, D., Schmit, J. L., & Vesper, N. (1999). Going to college: How social, economic, and educational factors influence the decisions students make Baltimore: Johns Hopkins Univ ersity Press. Hossler, D., & Stage, F. K. (1992). Family and high school experience influences on the postsecondary educational plans of ninth grade students. American Educational Research Journal, 29 (2), 425 451. Hsiao, K. P. (1992). First generation c ollege students Los Angeles, CA: ERIC Clearinghouse for Junior Colleges, UCLA. Hurtado, M. T. (1997). The influence of parents and siblings on late adolescents' academic achievement and college attendance (Doctoral dissertation). Available from ProQuest Dissertations & Theses database. (AAT 9816663)
206 Hurtado, S., & Carter, D. (1997). Effects of college transition and perceptions of the Sociology of Education, 70 (4), 324 345. Hurtado, S., Car ter, D. F., & Spuler, A. (1996). Latino student transition to college: Assessing difficulties and factors in successful college adjustment. Research in Higher Education, 37, 135 157. Hurtado, S., Inkelas, K. K., Briggs, C., & Rhee, B. S. (1997). Differenc es in college access and choice among racial/ethnic groups: Identifying continuing barriers. Research in Higher Education, 38 (1), 43. Hurtado Ortiz, M. T., & Gauvain, M. (2007). Postsecondary education among Mexican American youth: Contributions of paren ts, siblings, acculturation, and generational status. Hispanic Journal of Behavioral Sciences, 29 (2), 181 191. Ingels, S. J., & Dalton, B. W. (2008). Trends among high school seniors, 1972 2004 (NCES 2008 320). Washington, D.C.: U.S. Department of Educa tion, National Center for Education Statistics. Inkelas, K. K., & McCarron, G. (2006). The gap between educational aspirations and attainment for first generation college students and the role of parental involvement. Journal of College Student Development 47 (5), 534 549. Institute for H igher Education Policy. (2005). The investment payoff: A 50 state analysis of the public and private benefits of higher education. Washington DC: Author. Ishitani, T. T. (2003). A longitudinal approach to assessing attri tion behavior among first generation students: Time varying effects of pre college characteristics. Research in Higher Education, 44 (4), 433 449. Julian, T., & Kominski, R. (2011). Education and synthetic work life earnings estimates (ACS 14). Washington DC: American community survey reports. Retrieved from http://www.cens us.gov/prod/2011pubs/acs 14.pdf Jun, A., & Colyar, A. (2002). Parental guidance suggested: Family involvement in college preparation programs. In L. S. Hagedorn & W. G. Tierney (Eds. ), Increasing access to college: extending possibilities for all students (pp. 195 215). Albany: State University of New York Press. Karen, D. (2002). Changes in access to higher education in the United States: 19801992. Sociology of Education, 75 191 2 10. Kao, G., & Thompson, J. S. (2003). Racial and ethnic stratification in educational achievement and attainment. Annual Review of Sociology, 29 417 442.
207 Kao, G., & Tienda, M. (1998). Educational aspirations of minority youth. American Journal of Edu cation, 106 (3), 349. Kim, D. (2004). The effects of financial aid on students' college choice: Differences by racial groups. Research in Higher Education, 45 (1), 43 70. Kincheloe, J. L. (2004). Why a book on urban education? In S. Steinberg & J. Kincheloe (Eds.), 19 urban questions: Teaching in the city (1 32). ). New York, NY: Peter Lang Publishing. Kincheloe, J. L. (2010). Why a book on urban education? In S. Steinberg (Ed), 19 urban questions: Teaching in the city (2nd ed., pp. 1 28). New York, NY: Pe ter Lang Publishing. King, J. E. (2006). Gender equity in higher education: 2006. Washington, D.C.: American Council on Education Center for Policy Analysis Kirst, M. W., & Venezia, A. (2004). From high school to college: Improving opportunities for s uccess in postsecondary education San Francisco, CA: Jossey Bass. Kiyama, J. (2008). Funds of knowledge and college ideologies: Lived experiences among Mexican American families (Doctoral dissertation). Available from ProQuest Dissertations & Theses data base. (AAT 3303779) Knapp, G., Kelly Reid, J.E., Whitmore, & R.W., Miller, E. (2007). Enrollment in Postsecondar y Institutions, Fall 2005; Graduation Rates, 1999 and 2002 Cohorts; and Financial Statistics, Fiscal Year 2005 (NCES 2007 154). Washington, DC : U.S. Department of Education, National Center for Education Statistics. Kohler, A.D., & Lazarin, M. (2007). Hispanic education in the United States. National Council of LaRaza, 8, 1 16. Kotler, P., & Fox, K. F. A. (1985). Strategic marketing for educa tional institutions Englewood Cliffs, NJ: Prentice Hall. Kyriakides, L. (2005). Extending the Comprehensive Model of Educational Effectiveness by an Empirical Investigation. School Effectiveness and School Improvement, 16 (2), 103 152. Kuh, G.D., Cruce, T.M., Shoup, R., Kinzie, J., & Gonyea, 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. Kurlaender, M. (2006). Choosing community college: Factors affe cting Latino college choice. New Directions for Community Colleges, 2006 (133), 7 16.
208 Laden, B. V. (2004). Serving emerging majority students. New Directions for Community Colleges, 127 5 19. doi: 10.1002/cc.160 Lee, C. (2004). Racial segregation and e ducational outcomes in metropolitan Boston. Cambridge, MA: The Civil Rights Project at Harvard University. Li, Y., Sheely, M.C. II, & Whalen, D.F. (2005). Contributors to residence hall student retention: Why do students choose to leave or stay? Journal of Colleges & University Student Housing, 33 (2), 28 36. Litten, L. H. (1982). Different strokes in the applicant pool: Some refinements in a model of student college choice. Journal of Higher Education, 53 (4), 383 402. Loeb, S., Rouse, C., & Shorris, A. (2007). Introducing the issue. In authors names (Eds.), The future of children: Excellence in the classroom ( 3 14 ) Princeton Brooking, 17 (1). Lohfink, M., & Paulsen, M. B. (2005). Comparing the determinants of persistence for first generation and continuing generation students. Journal of College Student Development, 46 (4), 409 428. doi: 10.1353/csd.2005.0040 Lopez, G. (2001). The value of hard work: Lessons on parent involvement from an (im)migrant household. Harvard Educational Review, 71 (3), 41 6 438. Lowman, J., & Elliott, M. (2010). A multilevel model of educational expectations of secondary school students in the United States. Social Psychology of Education, 13 (1), 77 110. Luis Brown, D. (1967). Waves of dsheeecolonization: discourses of r ace and hemispheric citizenship in Cuba, Mexico, and the United States Durham: Duke University. Mack, K. (2011, June 27). Emanuel, Duncan praise federal DREAM Act in advance of Durbin's Senate hearing, Chicago Tribune Retrieved from http://articles.chic agotribune.com/2011 06 27/news/chi emanuel duncan praisefederal dream act in advance of durbins senate hearing 20110627_1_dream actcitizenship for college students citizenship path Mau, W. C., & Bikos, L. H. (2000). Educational and vocational aspirations of minority and female students: A longitudinal study. Journal of Counseling & Development, 78 (2), 186 194. McDonough, P. M. (1997). Choosing colleges: How soci al class and schools structure opportunity Albany: State University of New York Press.
209 Mc Donough, P.M. (2004). Counseling matters: Knowledge, assistance, and organizational commitment in college preparation. In W.G. Tierney, Z.B. Corwin, & J.E. Colyar (Eds.), Preparing for college: Nine elements of effective outreach. Albany, NY: State Univers ity of New York Press. McDonough, P M. (2005). Counseling matters: Knowledge, assis tance, and organizational commitment in college preparation. In W. G. Tiemey, Z. B. Corwin, & J. E. Colyar (Eds.), Preparing for college: Nine elements of effective outre ach (pp. 69 87). Albany: State University of New York Press. Mello, Z. R. (2009). Racial/ethnic group and socioeconomic status variation in educational and occupational expectations from adolescence to adulthood. Journal of Applied Developmental Psycholog y, 30 (4), 494 504. Mendez, A. R. (1994). Cubans in America Minneapolis: Lerner Publications Co. Mendez, J. (2003). How students understand financial aid: A qualitative study of the college choice process (Doctoral dissertation). Available from ProQuest Dissertations & Theses database. (AAT 3110533) Meredith, D. (2008). Carolina Covenant: Low SES, first generation college students navigation of higher education (Doctoral dissertation). Available from ProQuest Dissertations & Theses database. (AAT 33238 41) Millimet, D. (2001). bias? Retrieved February 4, 2010 from http://www.stata.com/support/faqs/stat/bias.html Monk, D (2007). Recruiting and retaining high quality teachers in rural areas. Future Child 17 (1), 155 74. Morse, A. (2013). In state tuition and unauthorized immigrant students Immigrant Policy Project, National Conference of State Legislatures. Museus, S.D. (2011). Generating ethnic minority success (GEMS): A qualitative analysis of high per forming institutions. Journal of Diversity in Higher Education, 4 (3), 147 162. National Association of Latino Elected and Appointed Officials (NALO) (2010). Report from the 6 th Annual Summit on the State of Latino Education September 27 29, 2010. Nav arrette, R., Jr. (2010, August). Dream Act a needed step in immigration reform. SanFrancisco Chronicle Retrieved from http://www.sfgate.com/cgibin/article.cgi?f=/c/a/2010/08/17/EDRG1EV9I8.DTL
210 NCHS. (2012). Health, United States, 2011: WIth special f eature on socioeconomic stauts and health Hyattsville, MD: National Center for Health Statistics. Noguera, P. (2003). City schools and the American d ream: Reclaiming the promise of public education. New York, NY: Teachers College Press. N ora, A., & Ca brera, A. F. (1992). Measuring program outcomes: What impacts are important to assess and what impacts are possible to measure? In Synthesis of Major Themes and Commissioned Papers Prepared for the Conference (Washington, D.C., September 30, 1992) (pp. 85 102). Westat Inc.: Rockville, MD. Nora, A., & Rendon, L. I. (1990). Determinants of predisposition to transfer among community college students: A structural model. Research in Higher Education,31 (3), 235 255. Nunez, A., & Cuccaro Alamin, S. (1998). Firs t generation students: Undergraduates whose parents never enrolled in postsecondary education (NCES Publication No.98082) Washington, DC: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics. O 'Connor, N., Hammack, F. M., & Scott, M. A. (2010). Social capital, financial knowledge, and Hispanic student college choices. Research in Higher Education 51 (3), 195 219. doi: 10.1007/s11162 009 9153 8 Olive, T. (2008). Desire for higher education in fir st generation Hispanic college students enrolled in an academic support program: A phenomenological analysis. Journal of Phenomenological Psychology, 39 (1), 81 110. Oliverez, P. (2006). Ready but restricted: An examination of the challenges of college ac cess and financial aid for college ready undocumented students in the United States (Doctoral dissertation). Available from ProQuest Dissertations & Theses database. (AAT 3257819) Park, S., Umbach, P. D., Padgett, R. D., Wells, R. S., & Seifert, T. A. (20 11). Why do more women than men want to earn a four year degree?: Exploring the effects of gender, social origin, and social capital on educational expectations. The Journal of Higher Education, 82 (1), 1 32. Pascarella, E. T., Pierson, C. T., Wolniak, G C., & Terenzini, P. T. (2004). First generation college students: Additional evidence on college experiences and outcomes. Journal of Higher Education, 75 (3), 249 284. Passel, J.S., & Cohn, D. U.S. Unauthorized Immigration Flows are down sharply since m id Decade ( Washington, D.C. Pew Hispanic Center, 2010), http://pewhispanic.org/files/reports/126.pdf
211 Paulsen, M. B. (1990). College choice: Understanding student enrollment behavior. ASHE ER IC Higher Education Report No. 6. Washington DC: Association for the Study of Higher Education, Eric Clearinghouse on Higher Education, George Washington University School of Education and Human Development. Penrose, A. M. (2002). Academic literacy perce ptions and performance: Comparing first generation and continuing generation college students. Research in the Teaching of English, 4 (36), 437 461. Perez, D. (2010, August 22). The American DREAM. TheHuffingtonPost.com. Retrieved from http://www.huffing tonpost.com/david Prez /the americandream_b_687225.html Prez, L. X. (1999). The interface of individual, structural, and cultural constructs in Latino parents' effort to support their children in planning for college (Doctoral dissertation). Available f rom ProQuest Dissertations & Theses database. (AAT 9952503) Prez, P. A. (2007). Social capital and chain migration: The college choice process of Chicana and Chicano community college, transfer and university students (Doctoral dissertation). Available f rom ProQuest Dissertations & Theses database. (AAT 3272322) Prez, P. A., & McDonough, P. M. (2008). Understanding Latina and Latino college choice: A social capital and chain migration analysis. Journal of Hispanic Higher Education, 7 (3), 249 265. Peng, C., Lee, K., & Ingersoll, G. (2002). An introduction to logistic regression analysis and Reporting. The Journal of Educational Research, 96 (1), 3 14. Pennock Romn, M. (1990). Test validity and l anguage background: A study of Hispanic American students at six universities New York: College Entrance Examination Board. Perna, L. (2000). Differences in the decision to attend college among African Americans, Hispanics, and Whites. Journal of Higher Education, 71 (2), 117 141. Perna, L. (2004). Impact of s tudent aid program design, operations, and marketing on the formation of family college going plans and resulting college going behaviors of potential students Boston, MA: The Education Resources Institute, Inc ( TERI). Perna, L. (2006). Studying college access and choice: A proposed conceptual model. In J. C. Smart (Ed.), Higher education: Handbook of theory of research (pp. 99 157). New York: Springe
212 Perna, L., Rowan Kenyon, H., Bell, A., Thomas, S., & Li, C. (2008). A typology of federal and state pro grams designed to promote college enrollment. The Journal of Higher Education, 79 (3), 243 267. Perna, L., & Titus, M. A. (2005). The relationship between parental involvement as social capital and college enrollment: An examination of racial/ethnic group differences. Journal of Higher Education, 76 (5), 485 518. Perna, L. W. (2008). Understanding high school students' willingness to borrow to pay college prices. Research in Higher Education, 49 (7), 589 606. Perreira, K. M., Harris, K.M., and Lee, D. (200 7). Immigrant youth in the Labor Market. Work Publications, 34 (1). Pew Hispanic Center. (2009). Between two worlds: How Young Latinos come of age in America Washington, DC: Author. Pew Hispanic Center. (2010). Hispanics of Mexican origin in the United States, 2008. Washington, DC: Author. Phinney, J., Dennis, J., & Chuateco, L. (2005). The role of motivation, parental support, and peer support in the academic success of ethnic minority first generation college students. Journal of College Student Dev elopment, 46 (3), 223 236. Plank, S. B., & Jordan, W. J. (2001). Effects of information, guidance, and actions on post secondary destinations: A study of talent loss. American Educational Research Journal, 38, 947 979. doi: 10.3102/00028312038004947 Punke tt, S. W., & Bmaca Gomz, M. Y. (2003). The r elationship between parenting, acculturation, and adolescent academics in Mexican origin immigrant families in Los Angeles. Hispanic Journal of Behavioral Sciences, 25 222 239. Population Reference Bureau (2 010). being in Numbers and Trends. Washington, DC: Population Reference Bureau. Portes, P. R. (1999). Social and psychological factors in the academic achievement of children immigrants: A cultural history puzzle. Ameri can Educational Research Journal 36, 489 507. Portes, A., & Rumbaut, R. G. (2006). Immigrant America: A portrait (3rd ed.). Berkeley: University of California Press. Ramirez, R.R., and de la Cruz, G.P. (2003). The Hispanic population in the United Stat es: March 2002: Population characteristics. Current Population Reports, P20 545. Washington, DC: US Census Bureau. http://www.census.gov/prod/2003pubs/p20 545.pdf
213 Riehl, R. J. (1994). The a cademic preparation, aspirations, and first year performance of first generation students. College and University, 70 (1), 14 19. Rios Aguilar, C., & Deil Amen, R. (2012). Beyond getting in the fitting in: An examination of social networks and professiona lly relevant social capital among Latina/o university students. Journal of Hispanic Higher Education, 11 (2), 179 196. doi : 10.1177/1538192711435555. Rong, X. L., & Grant, L. (1992). Ethnicity, generation, and school attainment of Asians, Hispanics, and Non Hispanic Whites. The Sociological Quarterly, 33 (4), 625 636. Rooney, G. (2008). Low income, first generation, Africa n American and Latino students' perceptions of influencing factors on their successful path to enrollment in a four year college (Doct oral dissertation). Available from ProQuest Dissertations & Theses database. (AAT 3307619) Sabogal, F., Marn, G., Otero Sabogal, R., Marn, B., & Perez Stable, E.J. (1987). Hispanic Jou rnal of Behavioral Sciences, 9 397 412 Saenz, V.B. & Ponjuan, L. (2009). The vanishing Latino male in higher education. Journal of Hispanic Higher Education, 8 (54), 54 90, doi : 10.1177/1538192708326995 Saenz, V. B., Hurtado, S., Barrera, D., Wolf, D., & Yeung, F. (2007). First in my family: A profile of first generation college students at four year institutions since 1971 Los Angeles: Higher Education Research Institute, UCLA. Santiago, D. (2007). tudents. Washington, DC: Excelencia in Education. Retrieved from http://www.edexcelencia.org/research/voces profile todays latino collegestudents Santiago, D. (2008). The condition of Latinos in education: 2008 factbook Washington, DC: Excelencia in Education. Retrieved from http://www.edexcelencia.org/research/conditions latinos education 2008 factbook Santiago, D. (2011). Roadmap for ensuring America's fut ure by increasing Latino college completion. Washington, DC: Excelencia in Education. Retrieved from http://www.edexcelencia.org/initiatives/EAF/Roadmap Santiago, D., & Cunningham, A. F. (2005). How Latino students pay for college: Patterns of financial aid in 2003 04. Washington, DC: Excelencia in Education, Institute for Higher Education Policy. Retrieved from http://www.edexcelencia.org/research/how latino students pay college patternsfinancial aid 2003 04
214 Saunders, M., & Serna, I. (2004). Making coll ege happen: The college experiences of first generation Latino students. Journal of Hispanic Higher Education, 3 (2), 146 163. Sayrs, L. (1989). Pooled time series analysis Newbury Park, CA: Sage Publications, Inc. Schaller, A., Rocha, L.O., & Barshin ger, D. (2007). Maternal attitudes and parent own low levels of education. Early Childhood Education Journal, 34 (5), 351 356. Smith, M. J., & Fleming, M. K. (2006). African American parents in the search stage of college choice: Unintentional contributions to the female to male college enrollment gap. Urban Education, 41 (1), 71 100. doi: 10.1177/0042085905282255 Snyder, T. D., & Dillow, S. A. (2009). Digest of education stat istics Washington, DC: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics. Retrieved from http://nces.ed.gov/programs/digest/2009menu_tables.asp Solorzano, D.G., Villalpando, O., & Oseguera, L. (2005). Educational inequities and Latino/o undergraduate students in the United States: A critical race analysis of their educational progress. Journal of Hispanic Higher Education, 4, 272 294. Stanton Salazar, R. D. (2004). Social capital among working class m inority students. In M. A. Gibson, P. Gndara, & J. P. Koyama (Eds.), School Connections: U.S. Mexican Youth, Peers, and School Achievement. New York: Teachers College Press, Columbia University. Stanton Salazar, R. D., & Dornbusch, S. M. (1995). Social c apital and the reproduction of inequality: Information networks among Mexican origin high school students. S ociology of Education, 68 (2), 116 135. Steinberg, M. P., Piraino, P., & Haveman, R. (2008). Access to higher education: Exploring the variation in Pell grant prevalence among US colleges and universities. The Review of Higher Education, 32 (2), 235 270. St. John, E.P. (1990). Price response in persistence decisions: An a nalysis of the h igh s chool and b eyond senior cohort. Research in Higher Educat ion, 31(4), 387 403 Strayhorn, T. L. (2006). Factors influencing the academic achievement of first generation college students. NASPA, 43 (4), 82 111.
215 Strayhorn, T. L. (2009). Accessing and analyzing national databases. In T. J. Kowalski & T. J. L. II (E ds.), Handbook of data based decision making in education (pp. 105 122). New York, NY: Routledge Surez Orozco, C., & Surez Orozco, M. (1995). Transformations: Immigration, family life, and achievement motivation among Latino adolescents Stanford Univers ity Press. Swail, W. S., Cabrera, A. F., Lee, C., & Williams, A. (200 5a). From middle school to the workforce: Latino students in the educational pipeline. Latino students & the educational pipeline, Part I. Stafford, VA: Educational Policy Institute. S wail, W. S., Cabrera, A. F., Lee, C., & Williams, A. ( 2005b). Latino high school and baccalaureate graduates: A comparison. Latino students and the educational pipeline, Part II Stafford, VA: Educational Policy Institute. Swail, W. S., Cabrera, A. F., L ee, C., & Williams, A. (200 5c). Pathways to the bachelor's degree for Latino students. Latino students & the educational pipeline: Part III. Stafford, VA: Educational Policy Institute. Snyder, T.D., & Dillow, S.A. (2011). Digest of Education Statistics 20 11. U.S. Department of Education, National Center for Education Statistics 2012. Talavera Bustillos, V. H. (1998). Chicana college choice and resistance: An exploratory study of first generation Chicana college students (Doctoral dissertation). Available from ProQuest Dissertations & Theses database. (AAT 9906220) Terenzini, P. T., & Pascarella, E. T. (1998). Studying college students in the 21st century: Meeting new challenges. The Review of Higher Education, 21 (2), 151 165. Terenzini, P. T., Rendon, L I., Upcraft, M. L., Millar, S. B., Allison, K. W., Gregg, P. L., & Jalomo, R. (1994). The transition to college: Diverse students, diverse stories. Research in Higher Education, 35 (1), 57 73. Terenzini, P., Springer, L., Yaeger, P., Pascarella, E., & No ra, A. (1996). Firstgeneration college students: Characteristics, experiences, and cognitive development. Research in Higher Education, 37 (1), 1 22. Thayer, P. B. (2000). Retaining first generation and low income students. Opportunity Outlook, 2 2 8. Thomas, J.F. (1963). Cuban refugee program. Welfare in Review, 1 (3). Thomas, R. S. (1998, April). Black and Latino college enrollment: Effects of background, high school preparation, family and peer influence, and financial aid. Paper
216 presented at the A nnual Meeting of the American Educational Research Association, San Diego, CA. Retrieved from ERIC database. (ED420253) Thomas, S., & Heck, R. (2001). Analysis of large scale secondary data in higher education research: Potential perils associated with co mplex sampling designs. Research in Higher Education, 42 (5), 517 540. Tierney, W. G. (1999). Models of minority college going an d retention: Cultural integrity versus cultural suicide. The Journal of Negro Education, 68 (1), 80 91. Tierney, W. G., & Hagedo rn, L. S. (2002). Increasing access to college: extending possibilities for all students Albany: State University of New York Press. Tierney, W., & Venegas, K. (2006). Fictive kin and social capital. American Behavioral Scientist, 49 (12), 1687 1702. T into, V (1994). Learning college: Rethinking the causes of student attrition ( 2 nd ed.). Chicago, IL: University of Chicago Press. Titus, M.A. (2007). Detecting selection bias, using propensity score matching, and estimating treatment effects: An applica degree. Research in Higher Education, 48(4), 487 521 Torres, V. (2006). A mixed method study testing data model fit of a retention model for Latino students at urban universities. Journal of College Student Develo pment, 47 (3), 299 318. Torres, V., & Hernandez, E. (2009). Influence of an identified advisor mentor on urban Journal of college student retention, 11 141 160. Trotter, A. (2001). Report highlights progress, inequit y, and first generation college students. Education Week, 20 (39), 29 30. Trusty, J., Robinson, C. R., Plata, M., & Ng, K. M. (2000). Effects of gender, socioeconomic status, and early academic performance on postsecondary educational choice. Journal of C ounseling & Development, 78 (4), 463 472. U.S. Bureau of Labor Statistics, Office of Occupational Statistics and Employment Projections. (2009). Occupational Outlook Handbook, 2010 11 Edition Table 1. Retrieved from http://www.bls.gov/oco/oco2003.htm U.S. Bureau of Labor Statistics (2012). Economic News Release [News release]. Retrieved March 21, 2012 from http://www.bls.gov/news.release/emps it.t04.htm U.S. Census Bureau. (2007). Hispanic Americans: Census facts Retrieved from http://www.infoplease.com/spot/hhmcensus1.html
217 U.S. Census Bureau. (2009). American Community Survey. Retrieved from www.census.gov/acs/www/data_documentation/2009_release/ U.S. Census Bureau. (2012). Statistical Abstract of the United States: 2012 Washington, DC: Author. Retrieved from http://www.census.gov/compendia/statab/2012/tables/12s0300.pdf U.S. Department of Education, Institute of Education Sciences, & National Center for Education Statistics. (2010). Digest of Education Statistics: 2010 Retrieved from http://nces.ed.gov/Programs/digest/ U.S. Department of Education. National Center for Education Statistics. (2009). Digest of education statistics 2009 (Table 204). Washington, DC: U.S. Government Printing Office. U.S. Department of Education, Nat ional Center for Education Statistics. (2012). The Condition of Education 2012 (NCES 2012 045), Indicator 49 U.S. Department of Health & Human Services. (2013). Annual update of HHS povert y guidelines. Federal Register 78 (16), 5182 5183. Valencia, R. (2000). Inequalities a nd the schooling of minority students in Texas. Hispanic Journal of Behavioral Sciences 22, 445 459. Valenzuela, A. (1999). Subtractive schooling: U.S. Mexican you th and the politics of caring. Albany, NY: State University of New York Press Vela Gude L., Cavazos, J., Johnson, M.S., Fielding, C., Cavazos, A.C., Campos, L., students. Professional School Counseling, 12 (4), 272 279. Walpole, M. (2003). Socioeconomic status an d college: How SES affects college experiences and outcomes. The Review of Higher Education, 27 (1), 45 73. doi:10.1353/rhe.2003.0044 Warburton, E. C., Bugarin, R., & Nunez, A. M. (2001). Bridging the gap: Academic preparation and postsecondary success of first generation students (NCES No. 2001 153). Washington, D.C.: National Center for Education Statistics U.S. Government Printing Office. Wei, C.C., Berkner, L., He. S., Lew, S., Cominole, M., & Siegel, P. (2009). 2007 08 National Postsecondary Stude nt Aid Study (NPSAS : 08 ): Student Financial Aid Estimates for 2007 08: First Look (NCES 2009 166). National Center for Education Statistics, Institute of Education Sciences, U.S. Department of Education. Washington, D.C.
218 Wilson, P. M., & Wilson, J. R. (19 92). Environmental influences on adolescent educational aspirations. Youth and Society, 24 52 70. Wolf, L. (2007). A mixed method study of rural Iowa Latino high school students to determine barriers to access (Doctoral dissertation). Available from Pr oQuest Dissertations & Theses database. (AAT 3274892) York Anderson, D. C., & Bowman, S. L. (1991). Assessing the college knowledge of first generation and second generation college students. Journal of College Student Development, 32 (2), 116 122. Zara te, M. E., & Gallimore, R. (2005). Gender differences in factors leading to college enrollment: A longitudinal analysis of Latina and Latino students. Harvard Educational Review, 75 (4), 383 408. Zwick, R., & Sklar, J. G. (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), 439 464.
219 BIOGRAPHICAL SKETCH Griselda Flore s was born and raised in Pahokee, FL, a small town i n Palm Beach County. She attended the University of Florida where she earned her Bachelor of Arts in sociology and w omen studies Griselda went on to complet e her Master of Art in Teaching, majoring in secondary social s tudies e ducation and completed her i nternship at Clark Middle School. Immediately following, Griselda began her Ph.D. in h igher e ducation a dministration. Griselda is currently seeking employment at a postsecondary institution where s he can work with first generation and minority students.