Promoting or Perturbing Success

Abstract

Spanish

Using event history modeling, this study explored to what extent loans, grants, institutional aid, and work-study affect timing to first departure for Latino college students. The goal is to understand more about how aid promotes or perturbs success for Latino students as well as how those effects vary over time. Federal grants and targeted loans reduced likelihood of departure (attrition). Findings suggest that aid is a necessary but not sufficient condition for success.

Keywords

financial aid retention event history analysis new settlement state longitudinal methods

Although financial aid alone may not remove barriers to postsecondary attainment for students, studies have demonstrated that it can have an equalizing effect by encouraging preparation, promoting enrollment, and enabling students to focus on academics once enrolled (Flores & Chapa, 2009; Nora, 1990; St. John, Paulsen, & Carter, 2005). Although certainly not all Latinos¹ enrolled in postsecondary education come from a low-income background, their numbers are disproportionately high. Data from the 2008 American Community Survey show that Latinos aged 25 years and older have the lowest per capita income of any racial/ethnic group in the United States. This, coupled with the persistent attainment gap between Latinos and Whites at all levels of education (Aud, Fox, & KewalRamani, 2010) points to the need for greater understanding of the role, financial aid may play in promoting or perturbing postsecondary success among Latino students.

This study seeks to contribute to our knowledge of Latinos’ educational attainment by discerning the relationships between aid and first departure (i.e., not enrolling for credit) from postsecondary education. Specifically, this study explores the following question: “To what extent do types of aid—loans, grants, institutional aid, and work-study—affect the likelihood of departing from college among Latinos enrolled in public, baccalaureate degree-granting institutions in Indiana and how do these effects change over time?” In addition, this study adds to the growing body of work (e.g., Calcagno, Crosta, Bailey, & Jenkins, 2007; Chen & DesJardins, 2007; DesJardins, Ahlburg, & McCall, 1999) that models educational attainment as a longitudinal (rather than cross-sectional) process via event history analysis (EHA). The applied goal of this research is to identify periods in the educational trajectories of Latino students during which targeted aid might be most effective in promoting attainment.

Indiana: A New Settlement State

Indiana provides an important context for understanding the postsecondary pathways of Latino students. Although historically, relatively few Latinos have settled in the state, Indiana is among the states that have seen high rates of growth in Latino population more than the past 15 years. Whereas, Indiana has fewer Latinos than many other states (just more than 273,000 in 2005), its rates of Latino population growth have been higher than those of any of its Midwestern neighbors except Illinois—making Indiana a new settlement state. As in the new settlement states in the South (e.g., Georgia, North Carolina), Mexicans and Mexican Americans represent most of this growth (Fry, 2005a). Much of this increase has been concentrated in northwestern Indiana, near Chicago, and in Indianapolis. Between 2000 and 2004, Indianapolis had the fifth highest Latino growth rate of any metropolitan area in the country (Clark & Heet, 2006).

Whereas, the growth rate in Indiana among school-age Latino children, correspondingly, has outpaced that of Whites, considerable gaps exist between Whites and Latinos in educational outcomes. For example, nearly half of all Mexicans in Indiana did not complete high school, and all other racial and ethnic groups in Indiana were nearly twice as likely as Mexicans to have completed a postsecondary degree. Clark and Heet (2006) assert, “The single most important policy issue confronting Indiana vis-à-vis the growing Mexican population is in the realm of education. Mexican educational attainment suffers woefully compared to non-Mexican attainment” (p. 33). By 2020, it is projected that Latino students will comprise 22% of the total U.S. undergraduate population and 8% of the Indiana undergraduate population (Santiago & Cunningham, 2005).

With more and more Latino children entering the educational systems in Indiana and other new settlement states, it is imperative that we know more about the ways the education policy, including financial aid policy can help promote attainment by all students. This is particularly true in states like Indiana. Flores and Chappa (2009) note that the dispersion of Latinos into new settlement regions like Indiana may pose a challenge to policy makers and higher education systems as yet unfamiliar with serving Latino students.

Theoretical Framework

The study of student retention has been largely dominated by the student integration model (Spady, 1971; Tinto, 1975, 1982, 1988, 1993), which focuses on the ways that family background, precollege schooling, and individual attributes contribute to students’ initial institutional and goal commitments and affect integration into the social and academic spheres of the institution. A growing body of more recent evidence, however, suggests that factors such as mentoring (Torres, 2006), faculty interaction (Anaya & Cole, 2001), family support (Hernandez, 2000), and financial aid (Paulsen & St. John, 2002; St. John et al., 2005; Titus, 2006) contribute to retention differently for under-represented students than for White or students from higher-income groups. Research on the effects of campus climate has highlighted potential differences in factors affecting student retention along racial/ethnic and class lines (Nora & Cabrera, 1996; Tierney, 1992). Others (Hurtado, 2002; Hurtado & Carter, 1997; Miller & Sujitparapitaya, 2010; Nora & Cabrera, 1996; Rhee, 2008; Suarez-Balcazar, Orellana-Damacela, Portillo, Rowan, & Andrews-Guillen, 2003) have extended our understandings of the ways that student characteristics interact with institutional contexts and cultures through focusing on the contributions of campus climate to retention.

Rendón, Jalomo, and Nora (2000) offer a particularly poignant critique of integration models for the assimilationist approach, its focus on failure, its exclusion of historical and social factors, as well as its failure to consider systemic barriers to success. In sum, traditional models of persistence and student integration are not particularly well suited to conceptualizing the retention of Latino students (Oseguera, Locks, & Vega, 2009).

As Oseguera et al. (2009) note, more recent developments in retention theory are better suited to understanding the attainment of Latino students. The turn, toward more explicit consideration of contexts (e.g., St. John et al., 2005; Titus, 2006; Torres, 2006) is of particular relevance for this study. Building on prior work looking at the effects of environmental factors on persistence (Cabrera, Stampen, & Hansen, 1990), Cabrera, Castaneda, Nora, and Hengstler (1992) and Cabrera, Nora, and Castaneda (1993) developed and tested an integrated model that incorporates elements from both the integration and attrition models. Nora and Cabrera (1996) further developed this student adjustment model in testing the effects of prejudice and discrimination on the adjustment of under-represented students. The student adjustment model conceptualizes colleges as having academic and social domains in which students’ experiences can negatively or positively affect their cognitive and affective development, which in turn affects academic and intellectual development, commitment to degree attainment, and institutional commitment. Compared to earlier retention theory, the adjustment model incorporates greater consideration of student contexts as a major factor in persistence. This study utilizes this conceptual framework to guide selection of variables and statistical modeling, described in more detail below.

Method

Education researchers recognize implicitly the temporal dimension of attainment (e.g., Bean, 1985; Paulsen & St. John, 2002; Tinto, 1975), yet most employ cross-sectional analytic approaches and relatively few studies employ methods that incorporate time into their conceptual and analytic models. EHA is the longitudinal analysis of individuals’ or organizations’ experiences of events of interest over time (Allison, 1984). EHA explicitly incorporates time while also allowing for time-varying variables. For example, typical cross-sectional approaches to persistence may overlook variations in financial aid from one year to the next. Through EHA, we can incorporate real-world, year-to-year changes. The goal in this study is to build a temporal profile showing at what point in time a Latino student is most likely to depart (called a hazard profile in EHA) and how the aid might influence departure. (For a more detailed discussion of the use of EHA techniques in studying educational attainment, see DesJardins [2003]).

Data Sources and Sample

Because Latino students enrolled in Indiana’s postsecondary institutions are the population of interest, student-level data were derived from Indiana’s statewide longitudinal data system. These data come from the Indiana Commission for Higher Education statewide student information system (SIS) student-unit record database. SIS data are collected from all public postsecondary institutions in Indiana for enrollment-related transactions and represent the universe of students enrolled in Indiana’s public postsecondary institutions. Institutional price data from the Integrated Postsecondary Education Data System (IPEDS) along with receipt of aid data from SIS were used to calculate the net price of attending college (total cost of attendance less total aid) for each student. Total price included tuition, room, board, fees, books, supplies, and other expenses as reported by the institutions. Total aid was calculated as the sum of all forms of aid received by the student: private, institutional, state, and federal.

Latinos who (a) were first-time, first-year entrants between 1999-2000 and 2005-2006; (b) earned six credits during their first academic year; and (c) matriculated in a baccalaureate degree program constituted the sample of interest for this study. During this period (1999-2005), 24,631 Latino students were enrolled throughout the Indiana public postsecondary institutions, with 11,469 in baccalaureate degree programs. Of these, 4,963 students met the selection criteria, constituting the effective sample. Table 1 provides additional detail on the sample.

Table 1.

Sample Characteristics

Variables	Number of students	Column %
Institution type
State universities	434	8.7
Regional universities	2,453	49.4
Urban university	387	7.8
Research universities	1,689	34.0
Applied for aid	3,882	78.0
Adjusted gross income
Below US$19,000	844	17.0
US$19,000 to US$41,999	1,307	26.3
US$42,000 to US$78,999	1,411	28.4
US$79,000 and more	1,400	28.2
Received aid	3,589	72.3
Gender
Female	2,199	44.3
Male	2,764	55.7
Combined SAT score
Low SAT (≤910)	2,124	42.9
Mid SAT (920-1020)	1,226	24.7
High SAT (≥1030)	1,606	32.4
College GPA
A	657	13.2
B	2,064	41.6
C or less	2,242	45.2
Student housing
On-campus housing	1,807	36.4
Off-campus housing	2,931	59.1
Lived with parents	185	3.7
Housing unknown or overseas	40	0.8
Major declared
No	1,003	20.2
Yes	3,960	79.8
Total	4,963	100.0

Note: GPA = grade point average.

Although it is important to understand the experiences of Latino students in community colleges in new settlement states, this study focuses on Latino students enrolled in baccalaureate degree programs. Partly because of the lack until 2004 of a community college statewide system in Indiana, most Latino students in the state were enrolled in baccalaureate degree programs during the study period, while about 24% had enrolled in a community college. Although this latter number is not insignificant, given the methodological rationale (discussed next) for conducting separate analyses on baccalaureate-granting institutions and community colleges, it is reasonable to focus first on the sector enrolling the greatest number of the study population.

The methodological reason for focusing on one sector over another relates to concerns about unobserved differences in students—in this case, aspirations and motivations. Students who began at baccalaureate degree-granting institutions, arguably, had different aspirations and motivations than students who began at community colleges. Moreover, students’ aspirations and motivations may be related to the role financial aid plays in promoting attainment. This issue of unobserved characteristics is a persistent concern in the study of financial aid (DesJardins et al., 1999; Dowd, 2006) and it can confound attempts to model the effects of aid over time. Lacking data regarding students’ degree aspirations, focusing on baccalaureate degree-seeking students provides some control for different student intentions. Future work by this author will consider the patterns and pathways of Latino students enrolled in community colleges.

Models and Method

Departure from public postsecondary education was the event of interest in this study. Students who did not earn credit during fall and spring terms were considered to have departed at the end of their last year of enrollment. The definition of departure used here did not focus on departure from a single institution but rather on departure from an entire state’s system of public postsecondary education. This definition is consistent with recent research on educational mobility and social stratification that recognizes the increasingly complex patterns of student enrollment (Adelman, 2006). Students were removed from the sample after departing, meaning that stopping-in or -out was not examined for the purposes of this study. Graduates were considered not to have experienced the event of interest.

As suggested by Singer and Willett (2003) modeling began with a baseline model that shows the hazard profile for first departure without considering any of the other variables (e.g., academic preparation, aid). This generates a profile against which subsequent models with additional control variables can be compared.

Time (t) is measured in academic years. As suggested by Allison (1984), in instances where time is measured in discrete units, it is appropriate to employ discrete-time methods. Equation 1 denotes the general form of the model where h(t_j) represents the hazard² rate at a discrete point in time (j), D represents the baseline hazard intercept parameter at time periods one through seven, and β₁ through β₅ represent the slope coefficients for the predictor variables. Time periods six and seven were combined because of the small number of students remaining in the population at those points. Hazard rates tell us the probability of a student departing at time t, given that the student has not already departed, controlling for all else. Equation 1 below shows the general form of discrete-time survival model.

logit h (t_{j}) = [α_{1} + D_{1} + α_{2} D_{2} + \dots + α_{7} D_{7}] + [β_{1} x_{1} + β_{2} x_{2} + β_{3} x_{3} + β_{4} x_{4} + β_{5} x_{5}]

The models controlled for factors posited by the student adjustment model to affect academic success including (a) student background (x₁); (b) academic preparation (x₂); (c) college experiences and context (x₃); (d) academic momentum (x₄); and (e) financial aid (x₅) (See Equation 1). See Table 2 for a listing of the variables included in the event history models. All models included a dichotomous indicator of whether a student had applied for aid. The inclusion of a dichotomous indicator of aid application may serve as a proxy variable to help control for omitted variables (Cellini, 2008).

Table 2.

Variables Included in the Event History Models

Student background	Academic preparation	Academic momentum	College enrollment characteristics	Financial aid^a
GenderIncome^aAge^a	High school rank percentileCombined SAT scoreHigh school diploma type	Credits attempted^a20 Credits in Year 1	Campus residence^aIn-state residentDeclared major^aDevelopmental education^aInstitutional type^aCollege GPA	Institutional aidState grantsFederal grantsPrivate grantsNeed-based loansNon-need-based loansWork-studyCost of attendanceApplied for aid^bReceived aid^b

Note: GPA = grade point average.

Time-varying explanatory variables.

Categorical indicator.

Table 3.

Selected Regression Results From Proportional Hazards Model Showing Effects of Differentiated Aid on Hazard of First Departure

Variables	Significance	Odds ratio
Total cost	**	1.03
Federal grants	*	0.99
State grants		1.00
Need-based loans	**	0.99
Non-need-based loans	****	0.99
Institutional aid		1.00
Private gift aid		1.00
Work-study		0.99
Applied for aid	****	0.77
Person periods	11,863
−2 log likelihood		12,722.64
Model chi-square		1,331.31
Correctly predicted		65.30

p < .10. **p < .05. ***p < .01. ****p < .001.

Limitations

Several limitations warrant being mentioned before discussing findings from the models. Prior research (e.g., Fry, 2005b) has shown significant differences in the educational trajectory of first-, second-, or third-generation Latinos, as one example. Generational status is not accounted for in this study because these data were not available at the student level. In addition, data were not available for independent institutions such as the University of Norte Dame and Butler University, which enroll a greater proportion of Latino students than many of their public peers. Therefore, the findings cannot speak to the attainment processes of Latino students who enrolled in such relatively elite colleges.

Findings

Figure 1 shows the likelihood of departure (labeled “First Departure Hazard” on the vertical axis) in each time period of the study for the baseline, time-constant, and time-varying models. The baseline model excluded all variables (e.g., academic preparation, financial aid) except time, providing a reference point against which to judge subsequent models. The time-constant model assumed that the effects of each variable did not vary with time, in other words, that the effect of US$1,000 in grant aid was the same in Year 1 and Year 5. The time-varying model allowed the effects of aid to vary in each time period. A higher point on the vertical axis indicates a greater hazard that a Latino student will depart at the end of the corresponding time period. For example, findings from the baseline model indicate the hazard of departure increased significantly after each additional year of enrollment except Year 6 and Year 7, which were not statistically different from the reference year, Year 1.

Figure 1.

Fitted hazard profiles for baseline, time-constant, and time-varying models

Comparing the baseline to the time-constant model which included controls for student background, academic preparation, financial aid, and college enrollment characteristics yields a noticeable difference (higher points in the figure indicate greater hazard of departing after each point in time). Although initial hazard rates were similar after Year 1 (near 28%), when we controlled for variables hypothesized to affect student departure we saw a downward shift in the profile. Most noticeably, the hazard of departure actually decreased after Year 2 of enrollment but then increased after Year 3, stabilized after Year 4, and finally peaked after Year 5. We see from the profile for the time-varying model that financial aid appeared to lower the hazard of departure after certain time periods. An omnibus hierarchical test (Jaccard, 2001) indicated that the time-varying model significantly improved model fit (at the .05 level of significance), confirming our visual findings and supporting the conclusion that the effects of US$1,000 in different forms of aid is not the same in all time periods.

Total cost, federal grants, and need- and non–need-based loans were all significantly related to hazard of departure, although in each case the effect was relatively modest. A US$1,000 increase in total costs was associated with a 3% increase in the hazard of departing in any year, holding all else constant. In contrast, a US$1,000 increase in any type of loan or federal grant was associated with a decreased hazard of departure, controlling for all else. Receipt of aid was associated with a decreased hazard of departure, as was applying for aid. The effect for this categorical indicator was greater than the effects for the amount of loans or federal grants.

Discussion

The Effects of Grants and Loans

Consistent with prior work in this area (Olivas, 1985; Santiago & Cunningham, 2005), this study found that Latino students were most reliant on federal sources of aid (i.e., Pell and Stafford). However, the overall contribution of grants and loans to promoting persistence among Latino students was modest. Federal grants reduced the hazard of departure overtime, although the effect did not vary across time. In other words, federal grants did not appear to play a more or a less important role in departure in any given year. Pell grants constituted the single largest source of any form of grant aid on average, suggesting that federal grant aid and, more specifically, Pell grants provided a moderate foundation of financial support for Latino students and had a positive effect on reducing departure.

Need-based loans were significantly related to decreased hazard of departure after Year 4 and non–need-based loans had a similar effect at the end of Year 2. In comparison to the role of federal grants, it appears that Latino students relied on loans at different points in time. Need-based loans appeared to provide a consistent measure of financial support, but they were particularly impactful in reducing the hazard of departure at the end of the fourth year. Non-need-based loans, particularly parental loans were important sources of aid early on, but their usage decreased as students remained enrolled.

Sticker Shock

The most consistent finding in this study was the relationship between total cost and departure. The combination of a consistently significant relationship between cost and attendance along with a weaker and less consistent relationship between financial aid and attendance is intriguing. It would be reasonable to assume that if total cost were found to have a significant effect on departure, total aid would have an equally strong effect yet this is not the case in this study. The effects of aid do extend beyond the pecuniary to the psychosocial, however. Research has demonstrated that price responsiveness varies by student characteristics (Heller, 1997). For example, students from the low-income groups may be more responsive to different forms of aid than their higher-income peers (Paulsen & St. John, 2002). Moreover, some research (St. John et al., 2005) suggests that price responsiveness may differ across racial and ethnic groups and that students may be more responsive to costs than to aid (St. John & Starkey, 1995).

Overall, this finding suggests that the larger effect for total cost relative to aid found in this study may be an example of aversion to higher costs (sticker shock) among Latino students. If this is the case, receiving financial aid alone may not offset this effect. Indeed, prior research (Nora, Barlow, & Crisp, 2006) suggests that Latino students, particularly those with low incomes, may be more responsive to costs than to offers of aid. This issue warrants additional research and may be a fruitful direction for future study.

Implications

Several implications for practice emerge from this work. First, Latino students who applied for and received aid were less likely to depart than their peers who had not applied for aid after controlling for costs of attendance. This finding, consistent with prior work (Kimura-Walsh, Yamamura, Griffin, & Allen, 2009; Oseguera et al., 2009), suggests that having access to the necessary information to apply for aid is associated with an increased likelihood of enrolling and persisting. Outreach work regarding such information by colleges and universities may, therefore, play an especially positive role in promoting attainment among Latino students.

Another implication of this study is that financial aid offices at postsecondary institutions should consider to what extent Latino students are relying on loans in each year of enrollment and how information and aid might be packaged in time-varying ways. Non–need-based loans (mostly parental loans) were important in this study in reducing the likelihood of departure after Year 2, whereas Stafford loans were most efficacious after Year 4. Loan counseling and follow-up conversations with Latino students receiving aid may be as important in Years 2 and 4 as they are at the initial point of enrollment. Moreover, it may be helpful for institutions to target their scholarships and grant dollars at key points in time to reduce loan burden and promote attainment.

Several strands for further research emerge from the findings of this study. In particular, future research might consider Pell grant and Stafford loan use year-by-year among Latino students in different contexts to illuminate the nuanced and temporal nature of financing college. Another intriguing finding is the reliance among Latino students on parental loans in the early years of enrollment. Additional research into the use of parental loans by students might help us better understand communal approaches to financing education among Latino students.

Conclusion

In the nature of research, this study raises as many questions as it answers. What does the early use of parental loans by students tell us about the ways in which Latinos draw on family for support? To what extent is this reliance on parental loans indicative of and interwoven with a broader web of familial support structures? What are the underlying reasons Latino students may be particularly reliant on Stafford loans in their fourth year of enrollment? Is this reliance indicative of positive momentum toward graduation? Why does cost emerge as a consistently negative factor affecting departure when various forms of financial aid exert a relatively modest positive affect in this study?

This EHA of Latino postsecondary enrollment looked at one component of that temporal process—financial aid. Pell grants as well as Stafford and parental loans emerge from this research warranting further exploration. The modest positive findings here for the effects of financial aid indicate that it plays a necessary but not sufficient role in encouraging persistence among Latino students. This suggests that for financial aid to be an effective tool to help remove barriers to attainment, it must be one among many tools and, moreover, that we need to better understand the role played by financial aid at different points in time for Latino students.

Footnotes

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

References

Adelman

(2006). The toolbox revisited: Paths to degree completion from high school through college. Washington, DC: U.S. Department of Education.

Allison

P. D.

(1984). Event history analysis: Regression for longitudinal event data. Newbury Park, CA: SAGE.

Anaya

Cole

D. G.

(2001). Latina/o student achievement: Exploring the influence of student-faculty interactions on college grades. Journal of College Student Development, 42(1), 3-14.

Aud

Fox

M. A.

KewalRamani

(2010). Status and trends in the education of racial and ethnic groups. Washington, DC: Institute of Education Sciences, National Center for Education Statistics.

Bean

J. P.

(1985). Interaction effects based on class level in an explanatory model of college student dropout syndrome. American Educational Research Journal, 22(1), 35-64.

Cabrera

A. F.

Castaneda

M. B.

Nora

Hengstler

(1992). The convergence between two theories of college persistence. Journal of Higher Education, 63(2), 143-164.

Cabrera

A. F.

Nora

Castaneda

M. B.

(1993). College persistence: Structural equations modeling test of an integrated model of student retention. Journal of Higher Education, 64(2), 123-139.

Cabrera

A. F.

Stampen

J. O.

Hansen

W. L.

(1990). Exploring the effects of ability to pay on persistence in college. Review of Higher Education, 13, 303-336.

Calcagno

J. C.

Crosta

Bailey

Jenkins

(2007). Does age of entrance affect community college completion probabilities? Evidence from a discrete-time hazard model. Educational Evaluation and Policy Analysis, 29, 218-235.

10.

Cellini

S. R.

(2008). Causal inference and omitted variable bias in financial aid research: Assessing solutions. Review of Higher Education, 31, 329-354.

11.

Chen

DesJardins

S. L.

(2007). Exploring the effects of financial aid on the gap in student dropout risks by income level. Research in Higher Education, 49(1), 1-18. Retrieved from http://dx.doi.org/10.1007/s11162-007-9060-9

12.

Clark

Heet

(2006). Connecting Mexico and the Hoosier Heartland: Policy briefing. Indianapolis, IN: Sagamore Institute for Policy Research.

13.

DesJardins

S. L.

(2003). Event history methods: Conceptual issues and an application to student departure from college. Higher Education: Handbook of Theory and Research, 18, 421-471.

14.

DesJardins

S. L.

Ahlburg

D. A.

McCall

B. P.

(1999). An event history model of student departure. Economics of Education Review, 18, 375-390.

15.

Dowd

(2006). A research agenda for the study of the effects of borrowing and the prospects of indebtedness on students’ college-going choices. Boston, MA: New England Resource Center for Higher Education, University of Massachusetts.

16.

Flores

S. M.

Chapa

(2009). Latino immigrant access to higher education in a bipolar context of reception. Journal of Hispanic Higher Education, 8(1), 90-109.

17.

Fry

(2005a). A statistical portrait of Hispanics at mid-decade. Washington, DC: Pew Hispanic Center.

18.

Fry

(2005b). Recent changes in the entry of Hispanic and White youth into college. Washington, DC: Pew Hispanic Center.

19.

Heller

D. E.

(1997). Student price response in higher education: An update to Leslie and Brinkman. Journal of Higher Education, 68, 624-659.

20.

Hernandez

J. C.

(2000). Understanding the retention of Latino college students. Journal of College Student Development, 41, 575-587.

21.

Hurtado

(2002). Creating a climate of inclusion: Understanding Latino/a college students. In Smith

W. A.

Altbach

P. G.

Lomotey

(Eds.), The racial crisis in American higher education: Continuing challenges (Rev. ed., pp. 121-158). Albany: State University of New York Press.

22.

Hurtado

Carter

D. F.

(1997). Effects of college transition and perceptions of the campus racial climate on Latino college students’ sense of belonging. Sociology of Education, 70, 324-345.

23.

Jaccard

(2001). Interaction effects in logistic regression. Newbury Park, CA: SAGE.

24.

Kimura-Walsh

Yamamura

E. K.

Griffin

K. A.

Allen

W. R.

(2009). Achieving the college dream? Journal of Hispanic Higher Education, 8, 298-315.

25.

Miller

Sujitparapitaya

(2010). Campus climate in the twenty-first century: Estimating perceptions of discrimination at a racially mixed institution, 1994-2006. New Directions for Institutional Research, 2010(145), 29-52.

26.

Nora

(1990). Campus-based aid programs as determinants of retention among Hispanic community college students. Journal of Higher Education, 61, 312-331.

27.

Nora

Barlow

Crisp

(2006). Examining the tangible and psychosocial benefits of financial aid with student access, engagement, and degree attainment. American Behavioral Scientist, 49, 1636.

28.

Nora

Cabrera

A. F.

(1996). The role of perceptions of prejudice and discrimination on the adjustment of minority students to college. Journal of Higher Education, 67(2), 119-148.

29.

Olivas

M. A.

(1985). Financial aid packaging policies: Access and ideology. The Journal of Higher Education, 56, 462-475.

30.

Oseguera

Locks

A. M.

Vega

I. I.

(2009). Increasing Latina/o students’ baccalaureate attainment. Journal of Hispanic Higher Education, 8(1), 23-53.

31.

Paulsen

M. B.

St. John

E. P.

(2002). Social class and college costs: Examining the financial nexus between college choice and persistence. Journal of Higher Education, 73, 189-236.

32.

Rendón

L. I.

Jalomo

R. E.

Nora

(2000). Theoretical considerations in the study of minority student retention in higher education. In Braxton

(Ed.), Reworking the student departure puzzle (pp. 127-156). Nashville, TN: Vanderbilt University Press.

33.

Rhee

B. S.

(2008). Institutional climate and student departure: A multinomial multilevel modeling approach. Review of Higher Education, 31, 161-183.

34.

Santiago

Cunningham

(2005). How Latinos students pay for college: Patterns of financial aid in 2003-04. Washington, DC: Excelencia in Education and the Institute for Higher Education Policy

35.

Singer

J. D.

Willett

J. B.

(2003). Applied longitudinal data analysis: Modeling change and event occurrence. Oxford, UK: Oxford University Press.

36.

Spady

W. G.

(1971). Dropouts from higher education: Toward an empirical model. Interchange, 2(3), 38-62.

37.

St. John

E. P.

Paulsen

M. B.

Carter

D. F.

(2005). Diversity, college costs, and postsecondary opportunity: An examination of the financial nexus between college choice and persistence for African Americans and Whites. Journal of Higher Education, 76, 545-569.

38.

St. John

E. P.

Starkey

J. B.

(1995). An alternative to net price: Assessing the influence of prices and subsidies on within-year persistence. Journal of Higher Education, 66, 156-186.

39.

Suarez-Balcazar

Orellana-Damacela

Portillo

Rowan

J. M.

Andrews-Guillen

(2003). Experiences of differential treatment among college students of color. Journal of Higher Education, 74, 428.

40.

Tierney

W. G.

(1992). An anthropological analysis of student participation in college. Journal of Higher Education, 63, 603-618.

41.

Tinto

(1975). Dropout from higher education: A theoretical synthesis of recent research. Review of Educational Research, 45(1), 89-125.

42.

Tinto

(1982). Limits of theory and practice in student attrition. The Journal of Higher Education, 53, 687-700.

43.

Tinto

(1988). Stages of student departure: Reflections on the longitudinal character of student leaving. Journal of Higher Education, 59, 438-455.

44.

Tinto

(1993). Leaving college: Rethinking the causes and cures of student attrition (2nd ed.). Chicago, IL: University of Chicago Press.

45.

Titus

M. A.

(2006). Understanding college degree completion of students with low socioeconomic status. Research in Higher Education, 47, 371-398.

46.

Torres

(2006). A mixed method study testing data-model fit of a retention model for Latino/a students at urban universities. Journal of College Student Development, 47, 299-317.