Student Stop Out and Retention at a Hispanic-Serving Institution in Southern California: The Role of Background,Academic,and Environmental Factors

Abstract

French

This study explores the demographic/background, academic, and environmental factors that predicted student retention after stopping out at a large Hispanic-serving institution in Southern California. The results show that (a) gender, (b) academic background, experience, readiness, and performance, and (c) personal and financial issues predicted retention. We interpret these findings in the context of existing literature and provide suggestions to help institutions develop intervention strategies to promote student retention and “servingness” at HSIs.

Keywords

higher education Latina(o)retention gender academic experience environmental factors Hispanic-Serving Institution

Nationwide, changing cultural perspectives on the necessity of a college degree have resulted in growing demands for colleges and universities. For instance, between 1970 and 2017, student enrollment in higher education increased by 132% and, by 2018 to 2019, 21.9 million students were enrolled in colleges and universities across the nation (Hanson, 2021b). Higher education is increasingly seen as a necessary and essential step toward a secure economic future. As research shows, a college degree results in a more competitive workforce and lowers rates of poverty, unemployment, and incarceration (Stephan et al., 2015). However, despite the increasingly large number of students who pursue higher education, student retention remains a significant challenge. Indeed, for first-time, full-time degree-seeking undergraduate students who enrolled in 4-year degree-granting institutions in fall 2017, the overall retention rate was 81% (National Center for Education Statistics [NCES], 2019). At the least selective public institutions (i.e., those with an open admissions policy), the retention rate was 63% (NCES, 2019). Retention rates are even lower at public Hispanic-Serving Institutions (HSIs) in California enrolling many first-generation Latinx students. For instance, at the university where this study was conducted, the 4-year retention rate for students who enrolled as first-time, full-time freshmen in 2016 was 60.2%, indicating that two out of five students dropped out before completing their degree. These statistics suggest that while HSIs strive to serve Latinx students and raise their academic achievement, retaining them through degree completion remains problematic.

Investigating the factors associated with student attrition and retention at HSIs is particularly important. For instance, while “the ideal HSI identity is connected to equitable graduation and persistence rates for Latinxs” (Garcia, Nuñez & Sansone, 2019, p. 747), it is unclear how well HSIs serve Latinx students with regard to academic outcomes (Garcia, Nuñez & Sansone, 2019). Latinx undergraduate students are particularly affected by low degree completion rates compared to other ethnic groups (Fry & Lopez, 2012). Specifically, only 21% of the Latinx population between the ages of 25 and 29 holds a bachelor’s or higher degree compared to 45% of Whites (Postsecondary National Policy Institute [PNPI], 2020). Furthermore, of the 36% of Latinx students aged 18 to 24 enrolled in college in 2017, only 20% were enrolled in public 4-year bachelor’s degree programs compared to 56% of White students (PNPI, 2020). A recent systematic review of the literature of academic outcomes at HSIs revealed that college completion rates for Latinx students are not necessarily better at HSIs than at non-HSIs (Garcia, Nuñez & Sansone, 2019). The Latinx population is estimated to grow to almost 30% of the total US population by 2050 (Pew Research Center, 2020). Identifying the factors that predict Latinx students’ attrition and retention within the context of higher education, and especially at HSIs, is thus vitally important in order to promote their academic achievement and upward social mobility. Within this context, it is also necessary to increase HSIs’ “servingness” (Garcia, Nuñez & Sansone, 2019, p. 745) and effectiveness.

In this study, we explore the demographic/background, academic, and environmental factors that explain why students enrolled in a large metropolitan HSI in Southern California stopped out and, if they returned, what differences characterized returning and non-returning students. Specifically, we asked the following research questions:

Research Question 1: Are there differences in the characteristics of students who returned and those who did not return to the university after stopping out?

Research Question 2: What demographic/background, academic, and environmental factors predict student retention after the temporary withdrawal?

Conceptual Framework and Related Literature

The conceptual model guiding the present study combines elements of Tinto’s (1993) Model of Student Integration and Nora’s Student Engagement Model (Nora et al., 2006). Tinto’s model (Tinto 1975, 1993, 2012), which has evolved throughout the years, is the most influential in the field of academic student retention, providing a heuristic and theoretical framework for understanding student behavior. According to Tinto (1993), the decision to drop out arises from a combination of student characteristics and the extent of their academic, environmental, and social integration in an institution. Specifically, the model posits that student retention is based on a longitudinal process of interactions between an individual with given attributes, skills, financial resources, prior educational experiences, dispositions, and integration with other members of the academic and social systems of the institution (Tinto, 1993). This person-centered model, however, has been argued to be problematic when applied to minority students as it does not account for the crucial role of contextual factors influencing student success. Thus, Nora’s Student Engagement Model (Nora et al., 2006) was developed to include the myriad of individual, institutional, and environmental factors that lead to retention and graduation for Latinx students. In particular, the model posits that (a) pre-college factors such as demographic characteristics, skills, and attitudes, (b) academic factors and student experiences in college, and (c) environmental factors external to the college context (which are particularly relevant for students enrolled in HSIs) work together to influence students’ decisions to remain enrolled in college. The following section provides an overview of these factors, especially those included in the study.

Background/Demographic Factors

Research has consistently found that students’ demographic characteristics (age, gender, ethnicity, SES, etc.) and the skills and attitudes they bring to college (high school preparation, internal motivation, and commitment) influence their academic achievement and persistence in higher education. For example, several large-scale studies have demonstrated, for Latinx students, that being female is associated with higher academic achievement and being male is related to academic underachievement (Crisp et al., 2015). Research has shown that Latinx men face more challenges than Latinx women with adjusting to the academic college culture because of familial obligations and financial constraints (Witkow et al., 2015), cultural and familial expectations to provide (Sáenz et al., 2017), issues with accessing academic support services (Sáenz et al., 2016), and limited sense of belonging (Museus et al., 2017). However, at HSIs, Garcia et al. (2018) challenged the notion of underachieving Latinx men, and found that these students develop a deeper sense of self with regard to their racial/ethnic identity and are overall more successful than similar students at non-HSIs.

Academic experience (also measured in age) has also been found to affect retention outcomes. For instance, while nearly 30% of college freshman drop out before their sophomore year nationally, older students and students more ahead in their career are more likely to remain enrolled (Hanson, 2021a). In their systematic review on the factors that support the academic success of Latinx students, Winterer et al. (2020) found that the more units a student had completed, the greater the likelihood of student success. Byrd and Macdonald (2005) speculate that older students, especially those at HSIs, have had work and family experiences that contribute to the development of skills that are essential for college persistence. Also, students who have completed more units have developed a stronger sense of belonging to the university and are more likely to act as self-advocate while navigating through the college system, thus increasing their likelihood of persistence. Research at HSIs confirms that older, non-traditional students may be more motivated to persist at institutions that serve them well and enhance their self-perceptions of academic ability and potential (Cuellar, 2014; Garcia, Nuñez & Sansone, 2019).

Related to this finding is that entering college with a deﬁnite major and choice of career increases the probability of persistence (Willcoxson & Wynder, 2010). Indeed, high educational aspirations toward a specific bachelor’s degree and commitment to the institution have been found to be strongly related to students’ decisions to remain enrolled in college, especially for Latinx students (Crisp et al., 2015). This finding suggests that students who have chosen a career path and declared their major might be more likely to be retained than students who are undecided about their career.

Academic Factors

Academic factors have been reported as the overarching predictor of persistence for Latinx students (Crisp et al., 2015). Specifically, students’ cumulative grade point average (GPA), one of the most prevalent dependent measures of academic success, correlates positively with student outcomes, including student retention and their likelihood to earn a certificate or degree (Crisp & Nora, 2010; Mooring & Mooring, 2016). Stewart et al. (2015) found that first semester college GPA was a significant predictor of persistence for first-time students from varied backgrounds. However, Nora and Cabrera (1996) found that first-year GPA was three times more important in college persistence for Latinx students than for their White peers.

Another important academic factor, especially for Latinx students, is academic readiness as measured by their need for remediation and developmental courses and performance in gateway English and mathematics courses. Enrollment in developmental coursework correlates negatively with student outcomes and students who have required remedial classes are heavily represented among students who drop out (Crisp & Delgado, 2014; LaSota & Zumeta, 2016; Stewart et al., 2015). It is possible that student perceptions and attitudes related to their remediation are likely to impact their dropout behavior. If students do not see the value in remediation and developmental courses, their likelihood to remain enrolled in college is drastically minimized. Similarly, successful completion of ﬁrst-term mathematics and English composition courses has been found to predict retention and graduation (Musoba & Krichevskiy, 2014). The authors speculate that Latinx students’ performance on gateway English and mathematics courses might inﬂuence their perceptions of whether they belong in college, promoting persistence or withdrawal. At HSIs, research has shown that culturally relevant developmental programs and gateway curricula that take into consideration Latinx students’ racial and cultural ways of knowing and learning mitigate the negative perceptions around remediation, increasing success in first-year courses and promoting their persistence and degree completion (Excelencia in Education, 2008).

Environmental Factors External to College

Studies have shown that environmental factors external to the college context also influence Latinx students’ decisions to remain enrolled (Arana et al., 2011; Crisp et al., 2015; Kruse et al., 2015). Specifically, single parent status, having dependent children, and working full time correlate negatively with student outcomes, as these issues pull students away from full social and academic integration on campus, impacting their sense of belonging, desire, and ability to remain in college (Wood et al., 2012). Nora et al. (1996) found that Hispanic female students who reported taking care of a family member were 83% more likely to drop out than similar peers without such commitments. This was possibly due to their lack of time to devote to college or the stigmatization of having a family, especially in the context of non-HSIs that enrolled few non-traditional students. Markle (2015) reported that female students felt “put down” or “patronized” by professors for their parental responsibilities. Arana et al. (2011) similarly demonstrated that Latinx students who had families and worked while attending college found it challenging to juggle work and academic schedules and connect with the campus community:

For nonpersisters, it was extremely difficult to handle both the responsibilities of home life and school life [. . .] those that struggled seemed to be encumbered by family crises, family responsibilities, and lack of motivation [. . .] For each nonpersister, working presented itself as the major hurdle. Without flexible and diverse college schedules, this meant that these hurdles became impassable [. . .] There was a clear consensus among those that discontinued college: The interaction between work and college schedules was the major obstacle (pp.243–246)

At the same time, the authors found that students who reported having supportive faculty and staff were able to overcome environmental barriers, especially on campuses with a significant number of Latinx students and a recognizable Latinx cultural heritage that promoted culturally validating experiences. These findings suggest that HSIs’ characteristics may mitigate some of the contextual challenges faced by minority students.

Another environmental factor reliably shown to affect student success and retention is financial pressure to pay for college. Strained personal finances correlate negatively with student outcomes and Latinx students’ grades and decisions to persist in college are often related to family income (Kruse et al., 2015). Students who worry about covering the cost of college may indeed perform more poorly and ultimately drop out. In turn, a plethora of studies have shown that ﬁnancial aid substantially boosts retention for low-income and minority students and works to narrow the socioeconomic graduation gap. This is particularly true at HSIs, which specifically define themselves as serving Latinx students by seeking competitive federal grants that support student services and financial aid programs (Santiago et al., 2016). In these contexts, financial aid has been shown to be strongly associated with higher GPA and retention, especially for students who start college at a community college and attend college part time in order to work and pay for tuition, a pattern that may limit their chances to graduate (Contreras & Contreras, 2015; Kruse et al., 2015; Musoba & Krichevskiy, 2014). Likewise, HSIs that expend more financial resources on their students experience higher college graduation rates (Nuñez et al., 2011).

Method

Data Collection and the Sample

This study was conducted in one of the colleges of a large metropolitan public HSI in Southern California. The student population at this school consists of 94% underrepresented minority students, 69% of whom identify as Latinx. Most of the students at this institution are first-generation college students (58%), and an average of 47% graduate in 6 years or less. The study’s data were drawn from those students who enrolled in the college during the spring or fall of 2017 but did not return the semester following enrollment (N = 868). The project was supported by the Dean of the college, the second author of this study, and was carried out by an academic advisor and a tenured faculty of the college, the third and first author, respectively. Students who did not return following the first semester were contacted to explore reasons for their withdrawal and to help them reenroll. The resources of academic advising were used to carry out the project, which lasted one calendar year. A confidential survey was used to collect information systematically. Email, four waves of mailings (one of which included an incentive), and phone calls were utilized in an attempt to reach every student. The information collected from the students included demographic/background information, academic progress (e.g., GPA, completed and attempted units, disqualification, etc.), and their reasons for leaving, which included environmental variables external to college. The response rate was more than adequate: 785 students (90.4%) responded after several months of continuous efforts. Of these, 287 students (36.6%) returned to school by fall of 2020, whereas 498 (63.4%) did not.

Analytic Framework

We first ran chi-square and t-tests to identify significant differences in/relationships between the demographic, academic, and environmental characteristics of students who returned and students who did not. Next, since there were 23 variables that could be possible predictors of student retention and many of these were correlated, we conducted a principal component analysis to reduce the number of the predictor variables. We finally used logistic regression to ascertain the effects of (1) demographic variables (gender, ethnicity, and major), (2) GPA, and (3) the five components from the principal component analysis on the likelihood of participants’ retention. All analyses were run using SPSS 26.0. Evaluation of the logistic regression models involved an analysis of the chi-square goodness of fit, the Nagelkerke R², and the percentage accuracy in classification. The regression coefﬁcients (β), standard errors, the Wald statistics, associated p-values, and odds ratios for each predictor were also examined to identify significant relationships.

Results

Differences Between Retained and Non-Retained Students

Table 1 shows the demographic information, GPA, completed and attempted units, other academic variables (whether students had received probation, disqualification, and different types of holds, attended remedial mathematics and English courses, as well as their college-level mathematics and English grade), and reported reasons for not returning for retained and non-retained students. Chi-square and t-tests indicate statistically significant differences between the two groups on almost all variables except for major, ethnicity, and reporting financial concerns for not returning. These results suggest that retained students were more likely to be female and to have declared their major. Retained students were evenly split between students who had entered the university as freshmen and transfer students, whereas non-persisters had predominantly begun as freshmen. Retained students also had higher GPAs, had attempted and completed more units, and had received fewer probations, disqualifications, and academic/financial holds. Students who returned were also more likely not to have required remediation and to have completed gateway English and mathematics courses. Interestingly, retained students were more likely than those not retained to have reported personal/family reasons for stopping out; however, students who did not return were more likely to have reported failed and incomplete courses, other academic issues, and attending community college as reasons for not returning. On the other hand, the two groups of students did not differ in reporting financial/financial aid issues as reasons for stopping out, suggesting that financial concerns were similar for both retained and non-retained students.

Table 1.

Differences Between Retained and Non-Retained Students: Demographic/Background Variables, GPA, Completed and Attempted Units, Other Academic Variables, and Reported Reasons for Not Returning.

	Students who returned	Students who did not return	χ² or t		Students who returned	Students who did not return	χ² or t
Gender				Major
Female	227 (79.1)	344 (69.1)	χ²(1) = 9.21, p = .002	Soc work	26 (9.1)	29 (5.8)	χ²(6) = 8.39, p = .211
Male	60 (20.9)	154 (30.9)		Nursing	8 (2.8)	32 (6.4)
				Comm Dis	24 (8.4)	34 (6.8)
Ethnicity				Kinesiology	88 (30.7)	150 (30.1)
Latinx	204 (71.1)	356 (71.5)	χ²(6) = 4.24, p = .645	Ch&Fam St	49 (17.1)	83 (16.7)
Asian	31 (10.8)	57 (11.4)		Crim just	54 (18.8)	103 (20.7)
White	19 (6.6)	20 (4)		Pub Health	38 (13.2)	67 (13.5)
AfAm	15 (5.2)	26 (5.2)
AI/AN	1 (0.3)	5 (1)		Major status
NH/PI	1 (0.3)	4 (.8)		Pre-major/undecided	40 (13.9)	136 (27.3)	χ²(1) = 18.72, p < .001
Not reported	16 (5.6)	30 (6)		In major	247 (86.1)	362 (72.7)
College entry
Freshmen	146 (50.9)	355 (71.3)	χ²(1) = 32.86, p < .001	GPA	2.69 (.51)	2.28 (.87)	t(782) = -8.04, p < .001
Transfer	141 (49.1)	143 (28.7)
Units	84.57 (45.37)	64.92 (49.8)	t(643) = -5.63, p < .001	Attempt units	96.43 (50.38)	77.36 (53.22)	t(623) = -5.00, p < .001
Probation				Disqualification
Yes	35 (12.2)	94 (18.9)	χ²(1) = 5.92, p < .05	Yes	16 (5.6)	98 (19.7)	χ²(1) = 29.18, p < .001
No	252 (87.8)	404 (81.1)		No	271 (94.4)	400 (80.3)
Hold in remedial classes				Financial hold
Yes	14 (4.9)	88 (17.7)	χ²(1) = 26.36, p < .001	Yes	7 (2.4)	77 (15.5)	χ²(1) = 32.31, p < .001
No	273 (95.1)	410 (82.3)		No	280 (97.6)	421 (84.5)
Advisement hold				GPA hold
Yes	37 (12.9)	136 (27.3)	χ²(1) = 22.03, p < .001	Yes	18 (6.3)	100 (20.1)	χ²(1) = 27.18, p < .001
No	250 (87.1)	362 (72.7)		No	269 (93.7)	398 (79.9)
Remedial math completion				Remedial English completion
Not required	143 (49.8)	146 (29.3)	χ²(2) = 44.42, p < .001	Not required	143 (49.8)	146 (29.3)	χ²(2) = 36.99, p < .001
Yes	124 (43.2)	249 (50)		Yes	143 (49.8)	336 (67.5)
No	20 (7)	103 (20.7)		No	1 (0.3)	16 (3.2)
College level math grade				College level English grade
No credit/not completed	46 (16)	169 (33.9)	χ²(4) = 47.90, p < .001	No credit/not completed	19 (6.6)	99 (19.9)	χ²(4) = 46.46, p < .001
A	15 (5.2)	49 (9.8)		A	33 (11.5)	55 (11.1)
B	39 (13.6)	68 (13.7)		B	64 (22.3)	124 (24.9)
C	43 (15)	66 (13.3)		C	27 (9.4)	72 (14.5)
Not required	144 (50.2)	146 (29.3)		Not required	144 (50.2)	147 (29.6)
Reason: personal/family				Reason: transfer to community college
Yes	128 (44.6)	90 (18.1)	χ²(1) = 63.87, p < .001	Yes	33 (11.5)	127 (25.5)	χ²(1) = 22.00, p < .001
No	159 (55.4)	408 (81.9)		No	254 (88.5)	371 (74.5)
Reason: failed classes/incomplete remedial classes				Reason: other academic
Yes	15 (5.2)	83 (16.7)	χ²(1) = 21.67, p < .001	Yes	35 (12.2)	111 (22.3)	χ²(1) = 12.25, p < .001
No	271 (94.8)	415 (83.3)		No	252 (87.8)	387 (77.7)
Reason: financial/financial aid
Yes	37 (12.9)	83 (16.7)	χ²(1) = 2.00, p = .157
No	250 (87.1)	415 (83.3)

Note. AI/AN = American Indian/Alaskan Native; NH/PI = native Hawaiian/Pacific Islander.

Principal Component Analysis

We ran a principal component analysis on the 23 variables reported in Table 1 to reduce their number. We first examined whether there were variables that did not have at least one moderate correlation (r ≥ .3.) with another one. This was the case for gender, ethnicity, and major, which probably measured something different than the other variables. These variables were removed from the principal component analysis but were retained and separately entered in the regression. Using a direct oblimin rotation, the principal component analysis produced six factors with eigenvalues greater than 1. An inspection of the rotated component matrix revealed a complex structure, that is, two components loaded on the same variable, GPA. Thus, GPA was removed from the principal component analysis but retained and separately entered in the logistic regression.

A second principal component analysis was run on the remaining 19 variables. The correlation matrix showed that all variables had at least one correlation coefficient greater than .3. The overall Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy was 0.82, which is considered “meritorious” on Kaiser’s (1974) classification of measure values. The Bartlett’s Test of Sphericity was statistically significant (p < .0005), indicating that the data was likely factorizable. This analysis revealed five factors with eigenvalues greater than one which explained 35.2%, 12.8%, 9.6%, 8.8%, and 6.4% (for a total of 72.8%) of the total variance. Visual inspection of the scree plot indicated that the five components should be retained, especially since a five-component solution was consistent with the different background, academic, and environmental variables that the survey was designed to measure, with strong loadings on “academic background” variables (Component 1), “academic issues” (Component 2), “failed classes/incomplete remedial classes” variables (Component 3), “advisement-related issues” (Component 4), and “personal and financial issues” (Component 5). Table 2 reports the variables in each component and the component loadings and communalities of the rotated solution.

Table 2.

Results of the Principal Component Analysis: Variables Included Within Each Component, Component Loadings (CL), and Communalities (COMM).

	CL	COMM		CL	COMM
Component 1: academic background			Component 3: failed classes/incomplete remedial classes
College Entry (freshman vs. transfer)	0.99	0.87	Reason: failed class/incomplete rem class	−0.87	0.84
Remedial English completion	−0.99	0.88	Hold in remedial classes	−0.87	0.84
College level English grade	0.92	0.79	Reason: community college	−0.80	0.63
College level math grade	0.84	0.83	Reason: community college	−0.80	0.63
Remedial math completion	−0.83	0.90	Component 4: advisement-related issues
Units	0.72	0.79	Advisement hold	0.90	0.81
Attempted units	0.69	0.71	Probation	0.86	0.77
Major status (pre- vs. undecided)	0.50	0.30	Probation	0.86	0.77
Major status (pre- vs. undecided)	0.50	0.30	Component 5: financial/personal issues
Component 2: academic issues			Reason: financial/financial aid issues	−0.77	0.65
Disqualification	−0.97	0.92	Reason: personal/family issues	−0.59	0.59
GPA hold	−0.96	0.91	Financial hold	−0.57	0.41
Reason: other academic issues	−0.51	0.41

Five scales were constructed by using the component analysis scores automatically calculated by SPSS. These are calculated by determining the regression weights, multiplying each participant response by the respective weights, and then summing the products. This resulting sum (the “component score”) was the score a participant achieved on that particular component.

Modeling the Likelihood of Participants’ Retention: Results From a Logistic Regression

In the next step, we used logistic regression to ascertain the effects of (1) demographic variables (gender, ethnicity, and major), (2) GPA, and (3) the five components from the principal component analysis (academic background, academic issues, failed classes/incomplete remedial classes, advisement-related issues, and financial/personal issues) on the likelihood of participants’ retention. The model included dummy variables for each ethnic group contrasted against the majority Latinx group, and for each major contrasted against the largest major, Kinesiology. The results are shown in Table 3. A test of this model against a constant-only model attained statistical signiﬁcance, χ²(19) = 209.79, p < .001. The model explained 32.8% (Nagelkerke R²) of the variance in student retention, indicating that the set of background variables reliably distinguished between students who did and did not return the semester following enrollment. Overall, 71% of students were successfully classiﬁed using 0.5 as the cutoff point.

Table 3.

Logistic Regression Predicting Student Likelihood of Retention Based on Demographic Variables (Gender, Ethnicity, and Major), GPA, and the Five Components From the Principal Component Analysis (Academic Background, Academic Issues, Failed/Incomplete Remedial Classes, Advisement-Related Issues, and Financial/Personal Issues).

Predictor	β	SE β	Wald’s χ²	df	p-Value	OR
Constant	−.52	.58	0.80	1	.371	0.59
Gender (base: female)	−.80	.22	13.45	1	.000	0.45
Ethnicity (base: latinx)			1.61	6	.952
Asian	−.15	.28	0.29	1	.589	0.86
No report	−.29	.39	0.53	1	.466	0.75
White	.02	.38	0.00	1	.957	1.02
Haw/PI	−.67	1.20	0.31	1	.578	0.51
AmInd/AN	.22	1.19	0.03	1	.851	1.25
AfAm	.25	.41	0.38	1	.537	1.29
Major (base: kinesiology)			10.87	6	.090
Criminal justice	−1.49	.60	6.23	1	.013	0.22
Child and family studies	−.55	.45	1.48	1	.223	0.57
Public health	−.60	.38	2.42	1	.120	0.55
Communication disorders	−.61	.40	2.35	1	.125	0.55
Social work	−.19	.40	0.23	1	.630	0.82
Nursing	−.87	.42	4.38	1	.036	0.42
GPA	.16	.17	0.82	1	.366	1.17
Academic background	.42	.11	14.29	1	.000	1.52
Academic issues	−.52	.16	26.16	1	.000	0.57
Failed class/incomp rem class	−.46	.12	15.46	1	.000	0.63
Advisement-related issues	.20	.11	3.61	1	.057	1.23
Personal/financial issues	−.60	.10	36.66	1	.000	0.55
Overall model evaluation		Wald’s χ2 = 209.79			df = 19	p = .000
		Nagelkerke R2 = .328
		71.0% Successful classification
Hosmer and Lemeshow test		Wald’s χ2 = 11.56			df = 8	p = .172

The Wald criterion indicated that personal and financial issues, χ²(1) = 36.66, p < .001, academic issues, χ²(1) = 26.16, p < .001, failed classes/incomplete remedial classes, χ²(1) = 15.46, p < .001, academic background, χ²(1) = 14.29, p < .001, and gender, χ²(1) = 13.45, p < .001 were the best predictors of retention. For each point increase on the personal and financial issues, academic issues, and failed classes/incomplete remedial classes components (where higher scores represented more reported problems or failed/incomplete classes), the odds of returning decreased approximately by half. Likewise, male students were approximately half as likely to return as females. On the other hand, an increase in one unit in the academic background component (where higher scores represented a declared major, entry as a transfer student, more completed/attempted units, no need of remediation, and completion of/no requirement of college-level English and mathematics courses) increased the odds of returning by 1.52. Finally, while major per se did not predict retention, the majors of Criminal Justice χ²(1) = 6.23, p < .05, and Nursing, χ²(1) = 4.38, p < .05 also made smaller but statistically significant contributions to the prediction of retention, the odds ratios of these variables suggesting that students in these majors were less likely to return than students from Kinesiology, the major that was used as base for all comparisons. Interestingly, GPA per se did not predict retention. We speculate that GPA alone was less predictive of retention than academic issues taken collectively because retained and non-retained students did not differ widely in this measure (the former had a B− average, while the latter a C+). Also, most variables within the academic issues and failed classes/incomplete remedial classes components were related to GPA, suggesting that lower grades did affect retention but they did so indirectly by resulting in other academic problems (i.e., probation, disqualification, etc.). Likewise, neither ethnicity nor advisement-related issues predicted retention.

Discussion

The findings from this study contribute to our understanding of the factors influencing student retention after a temporary withdrawal at a large, urban HSI in Southern California. As expected, background, academic and environmental variables interacted together to influence students’ decision to return to college. Among the demographic variables, only gender predicted retention, as female students were twice as likely to re-enroll in the university after stopping out than males. These results replicate previous findings of underachievement among college-attending Latinx men. Thus, despite having “the power to provide Latino males the opportunity to develop a deeper sense of self with regard to their racial/ethnic identity” (Garcia et al., 2018, p. 183), a variety of environmental circumstances may intercede, and HSIs may continue to struggle to promote social and academic integration and academic success among Latinx male students as compared to Latinx females.

Academic background and experience, which was measured by the academic background component, also predicted retention. Students who declared their major, attempted and completed more units, and entered as transfer students were more likely to return than students who had completed fewer units and not declared a major. Likewise, academic readiness, also measured by the academic background component based on students’ need of remediation and performance in/completion of gateway English and mathematics courses, also predicted retention. Overall, our results suggest that more seasoned students, students who are more prepared at college entry and who aspire toward a specific degree may develop greater academic integration, a stronger sense of belonging to the university, and more motivation to persist, especially at institutions that serve them well and enhance their self-perceptions of academic ability and potential (Cuellar, 2014; Garcia, Nuñez & Sansone, 2019).

Academic performance—measured by the academic issues and failed classes/incomplete remedial classes components—also predicted retention. Specifically, students who had been disqualified, had GPA holds, and had reported academic problems were less likely to return than students who had not experienced these issues. Likewise, students who reported failed classes or incomplete required remediation and who attended the community college were also less likely to return to the university after stopping out. Thus, it appears that academic factors may be the overarching predictors of persistence for Latinx students at HSIs (Crisp et al., 2015). Contreras and Contreras (2015) argue that Latinx students are less likely than their White peers to be college ready as a result of not taking the set of high school classes that prepare them for enrollment in 4-year institutions. This lack of academic preparation prevents them from completing remedial and gateway courses, delaying their exposure to the curriculum of their intended major, lowering motivation, and encouraging dropout behavior. As put by Contreras and Contreras (2015, pp. 152–154), despite “consider[ing] the concept of ‘Hispanic-serving’ central to their institutional identity,” “at many of these institutions, targeted efforts to raise Latino academic performance are unclear,” and institutions often fail to address the academic needs of “the critical mass of Latino students [they serve] and improve their academic outcomes.”

Likewise, HSIs may struggle to address the greater personal and financial needs experienced by Latinx students attending these institutions (Nuñez et al., 2011). In this study, personal and financial variables external to college, measured by the personal/financial issues component, interacted with background and academic factors to promote or reduce retention. In particular, students who reported personal/family issues and financial/financial aid concerns or holds were half as likely to re-enroll in the university after stopping out than students who did not experience these problems. Nonacademic issues related to external responsibilities such as family commitments have been shown to be the most perceptible barriers to pursuing a college degree for Latinx students (Crisp et al., 2015; Kruse et al., 2015). These issues pull students away from social and academic integration, affecting their ability to remain in college and pursue a degree. Likewise, financial concerns may lead students to work to pay for tuition, decreasing their academic performance and their chances for retention and graduation. As put by Contreras and Contreras (2015):

The low socioeconomic backgrounds of Latinx students, coupled with their debt averse approach to college (Cunningham & Santiago, 2008) have resulted in a large proportion of Latino students working greater than 20 hours a week (Gandara & Contreras, 2009). Working more than 20 hours a week influences the amount of time spent on studying, the ability to be engaged on their college campus, the likelihood of college departure, and lengthens the overall time to degree completion (Gandara & Contreras, 2009, p.156)

Practical Implications, Limitations, and Directions for Future Research

Since “the ideal HSI identity is connected to equitable graduation and persistence rates for Latinxs” (Garcia, Nuñez & Sansone, 2019, p. 747), our findings suggest that HSIs should make a specific effort, from early on, to assist at-risk students and alleviate the challenges documented by the students in this study. Garcia et al. (2018) have shown that college-attending Latinx men develop a deeper sense of self with regard to their racial/ethnic identity and are overall more successful at HSIs than similar students at non-HSIs. Thus, HSIs should make a more concerted effort to reach out to male students at the earliest stages of their academic career to increase their social and academic integration, their sense of belonging to, and overall success at the university. Since Latinx students attending HSIs have greater academic and financial needs than traditional students (Nuñez et al., 2011), HSIs should also consider having systems in place to provide academic tutoring and peer mentoring for students who require remediation, have not completed gateway courses, or have done poorly in their course work. Institutions serving large numbers of Latinx students should also consider having specific resources and programs to help students who struggle with personal and financial issues. For example, since childcare has been shown to interfere with student retention, HSIs could explore developing child care centers that would be easily accessible to students with childcare needs while attending classes. In addition, because ﬁnancial aid has been shown to substantially boost retention for low-income and minority students, HSIs should consider greater financial aid to promote success in their students’ academic pursuits.

Ultimately, the results of this study should inspire HSIs to reassess their identity and examine how to better serve their Latinx students. Aguilar-Smith (2021) argues that most HSIs, in their quest to secure federal funding, end up negatively portraying or even losing sight of the needs of the very same Latinx students who make them eligible for funding. Similarly, Contreras and Contreras (2015) contend that HSIs may use Latinx students as commodities to secure funding that finances university-wide projects that bring few benefits to Latinx students. It is therefore imperative that HSIs re-examine their Hispanic-serving identity and strive to improve both student access and success at their institutions. This means reexamining academic outcomes, student experiences, curricula, programs, practices, and policies that may influence their degree of servingness. In the words of Garcia, Ramirez, Patrón and Cristobal (2019):

As organizations that lack a historical, articulated identity for serving Latinx students, HSIs continue to grapple with uncertainty in their identity. Yet if HSIs are going to reach their full potential, there is a need to move from an ambiguous, federally constructed identity into a measurable construct that can be operationalized in research and practice . . . [and] contribute to tangible measures of “servingness” (p. 531)

While the present study advances the debate on factors that promote student retention at a large, metropolitan HSI, it has several limitations that should be considered when interpreting the results. First, this study was conducted at a large metropolitan university in Southern California and included students from a college where female students were in the majority. Our results might not be generalizable to other colleges with fewer women. Second, the survey that was used to collect information had some limitations. In particular, some items were not specific enough to capture the full context of the variable that was intended to measure. For example, students reported “personal issues” as reasons for stopping out, but it was not clear which specific personal problems prevented them from reenrolling. Another limitation of the survey is that it did not capture whether students felt accepted, valued, and part of the university, variables that have been shown to be strong predictors of retention, especially at HSIs (Davidson et al., 2009).

Future studies would be enhanced by including a wider array of HSIs in their analyses. The federal government narrowly defines HSIs as “nonprofit, degree-granting postsecondary institutions that enroll at least 25 percent undergraduate Latinx students” (Garcia, 2019, p. 1). The diversity of HSIs is vast, including public and private schools, rural and urban institutions, and faith-based colleges. Beyond the Hispanic student enrollment criteria, there is often a rich diversity of students attending HSIs. With different institutional approaches to “serving” diverse students, future research would be enhanced by comparing the contexts of servingness in relationship to retention.

Researchers should also be remined to carefully refine their data collection instruments in order to capture the full context of the variables they are intended to measure. Future studies would be strengthened by comparing the different HSIs in relationship to students’ academic success and the demonstrated commitment of the institution through their various initiatives designed to impact retention. Within this context, thoroughly exploring the complex interplay of background, academic, and environmental variables that influence students’ decisions to return to college after stopping out is critical. Particular attention should be focused on the wide range of contextual factors that influence persistence, academic integration, commitment to the institution, and the degree of “servingness” of HSIs.

Footnotes

Declaration of Conflicting Interests

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

Funding

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

ORCID iD

Simona Montanari

Author Biographies

Simona Montanari, Ph.D., is a Professor in the Department of Child and Family Studies at California State University, Los Angeles. She received a Ph.D. in Linguistics from the University of Southern California specializing on early trilingual development. Her research has focused on the emergence of diﬀerent language components (speech sound production, vocabulary, verbal morphology, narrative skills, pragmatics) in Spanish-speaking children growing up bilingually or trilingually in the United States. Dr. Montanari has also been involved in the creation of an Italian/English dual language program in the Glendale Uniﬁed School District (Los Angeles County) and has been studying its educational outcomes since its launch in 2009. She also works as a consultant on bilingual/dual language education for diﬀerent schools in diﬀerent world regions and has designed curricula for Italian/English, French/English and German/English dual language programs.

Ron Vogel has served as a faculty member and administrator for more than 43 years. He started his academic career as a part-time lecturer in 1975 and became a tenure-track faculty member in 1979 after earning his doctorate from University of Massachusetts, Amherst. He went through the academic ranks at the University of North Carolina, Charlotte in the Department of Criminal Justice (1980-1991) primarily teaching research and served as a department chair (1988-91). In 1991, he moved to California State University, Long Beach (CSULB) and continued his career as a professor and chair of the Department of Criminal Justice for another 10 years before moving into administration. After interim positions as an Associate Dean and Associate Vice President for research and graduate programs, Ron became the Dean of the College of Health and Human Services at CSULB from 2002-2008. He has published dozens of peer reviewed research articles, worked on several editorial boards and served as the editor of the tier 1 research journal Crime and Delinquency. In 2009, he became Provost at California State University, Dominguez Hills before taking the position as Associate Vice Chancellor for the CSU system in 2011. In 2015, he returned to campus life and is currently the Dean of the Rongxiang Xu College of Health and Human Services at Cal State LA.

Melinda Vasquez, MS, is currently an Academic Advisor within the College of Health and Human Services at California State University, Los Angeles. She has also taught Introdution to Higher Education classes and Criminal Justice to students attending California State University Los Angeles and East Los Angeles Community College.

References

Aguilar-Smith

(2021). Seeking to serve or $erve? Hispanic-serving institutions’ race-evasive pursuit of racialized funding. Journal of Hispanic Higher Education, 7(1), 1–15. https://doi.org/10.1177/23328584211057097

Arana

Castañeda-Sound

Blanchard

Aguilar

T. E.

(2011). Indicators of persistence for Hispanic undergraduate achievement: Toward an ecological model. Journal of Hispanic Higher Education, 10(3), 237–251. https://doi.org/10.1177/1538192711405058

Byrd

K. L.

Macdonald

(2005). Defining college readiness from the inside out: First-generation college student perspectives. Community College Review, 33, 22–37.

Contreras

F. E.

Contreras

G. J.

(2015). Raising the bar for Hispanic serving institutions: An analysis of college completion and success rates. Journal of Hispanic Higher Education, 14(2), 151–170. https://doi.org/10.1177/1538192715572892

Crisp

Delgado

(2014). The impact of developmental education on community college persistence and vertical transfer. Community College Review, 42(2), 99–117. https://doi.org/10.1177/0091552113516488

Crisp

Nora

(2010). Hispanic student success: Factors influencing the persistence and transfer decisions of Latino community college students enrolled in developmental education. Research in Higher Education, 51(2), 175–194. https://www-jstor-org-443.web.bisu.edu.cn/stable/40542370

Crisp

Taggart

Nora

(2015). Undergraduate Latina/o students: A systematic review of research identifying factors contributing to academic success outcomes. Review of Educational Research, 85(2), 249–274. https://doi.org/10.3102/0034654314551064.

Cuellar

(2014). The impact of Hispanic-Serving Institutions (HSIs), emerging HSIs, and non-HSIs on Latina/o academic self-concept. Review of Higher Education, 37, 499–530. https://doi.org/10.1353/rhe.2014.0032

Cunningham

Santiago

(2008). Student aversion to borrowing: Who borrows, who doesn’t? Institute for Higher Education Policy and Excelencia in Education. https://files.eric.ed.gov/fulltext/ED503684.pdf

10.

Davidson

W. B.

Beck

H. P.

Milligan

(2009). The college persistence questionnaire: Development and validation of an instrument that predicts student attrition. Journal of College Student Development, 50(4), 373–390. https://doi.org/10.1353/csd.0.0079

11.

Excelencia in Education. (2008). Modeling Hispanic Serving Institutions (HSIs): Campus practices that work for Latino students. Author. https://www.edexcelencia.org/research/issue-briefs/modeling-hispanic-serving-institutions-hsis-campus-practices-work-latino

12.

Fry

Lopez

M. H.

(2012). Hispanic student enrollments reach new highs in 2011: Now largest minority group on four-year college campuses. Pew Hispanic Center. https://www.pewhispanic.org/2012/08/20/hispanic-student-enrollments-reach-new-highs-in-2011/

13.

Gandara

Contreras

(2009). The Latino education crisis. Harvard University Press.

14.

Garcia

G. A.

(2019). Defining “servingness” at Hispanic-Serving Institutions (HSIs): Practical implications for HSI leaders. American Council on Education. https://www.equityinhighered.org/resources/ideas-and-insights/defining-servingness-at-hispanic-serving-institutions-hsis-practical-implications-for-hsi-leaders/

15.

Garcia

G. A.

Núñez

A. M.

Sansone

V. A.

(2019). Toward a multidimensional conceptual framework for understanding “servingness” in Hispanic-serving institutions: A synthesis of the research. Review of Educational Research, 89(5), 745–784. https://doi.org/10.3102/0034654319864591

16.

Garcia

G. A.

Patrón

O. E.

Ramirez

J. J.

Hudson

L. T.

(2018). Identity salience for Latino male collegians at Hispanic Serving Institutions (HSIs), emerging HSIs, and non-HSIs. The Journal of Hispanic Higher Education, 17(3), 171–186. https://doi.org/10.1177/1538192716661907

17.

Garcia

G. A.

Ramirez

J. J.

Patrón

O. E.

Cristobal

N. L.

(2019). Constructing an HSI organizational identity at three Hispanic-serving Institutions in the Midwest: Ideal versus current identity. The Journal of Higher Education, 90(4), 513–538. https://doi.org/10.1080/00221546.2018.1522198

18.

Hanson

(2021a). College dropout rates. Education Data Initiative. https://educationdata.org/college-dropout-rates

19.

Hanson

(2021b). College enrollment & student demographic statistics. Education Data Initiative. https://educationdata.org/college-enrollment-statistics

20.

Kaiser

H. F.

(1974). An index of factorial simplicity. Psychometrika, 39, 31–36. https://doi.org/10.1007/BF02291575

21.

Kruse

Starobin

S. S.

Chen

Baul

Santos Laanan

(2015). Impacts of intersection between social capital and finances on community college students’ pursuit of STEM degrees. Community College Journal of Research and Practice, 39(4), 324–343. https://doi.org/10.1080/10668926.2014.981893

22.

LaSota

Zumeta

(2016). What matters in increasing community college students’ upward transfer to the baccalaureate degree: Findings from the beginning postsecondary study 2003–2009. Research in Higher Education, 57(2), 152–189.

23.

Markle

(2015). Factors influencing persistence among nontraditional university students. Adult Education Quarterly, 65, 267–285. https://doi.org/10.1177/0741713615583085

24.

Mooring

R. D.

Mooring

S. R.

(2016). Predictors of timely baccalaureate attainment for underrepresented minority community college transfer students. Community College Journal of Research & Practice, 40(8), 681–694. https://doi.org/10.1080/10668926.2015.1070775

25.

Museus

S. D.

Saelua

(2017). The impact of culturally engaging campus environments on sense of belonging. Review of Higher Education: Journal of the Association for the Study of Higher Education, 40(2), 187–215. https://doi.org/10.1353/rhe.2017.0001

26.

Musoba

G. D.

Krichevskiy

(2014). Early coursework and college experience predictors of persistence at a Hispanic-serving institution. Journal of Hispanic Higher Education, 13, 48–62. https://doi.org/10.1177/1538192713513463

27.

National Center for Education Statistics. (2019). Digest of educational statistics 2019. Author. https://nces.ed.gov/programs/digest/d19/

28.

Nora

Barlow

Crisp

(2006). An assessment of Hispanic students in four-year institutions of higher education. In Castellanos

Gloria

A. M.

Kamimura

(Eds.) The Latina/o pathway to the Ph.D. (pp. 55–78). Stylus Publishing, LLC.

29.

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, 119–148. https://doi.org/10.1080/00221546.1996.11780253

30.

Nora

Cabrera

A. F.

Hagedon

L. S.

Pascarella

E. T.

(1996). Differential impacts of academic and social experiences on college-related behavioral outcomes across different ethnic and gender groups at four-year institutions. Research in Higher Education, 37, 427–451. https://doi.org/10.1007/BF01730109

31.

Nuñez

A. M.

Sparks

P. J.

Hernandez

E. A.

(2011). Latino access to community colleges and Hispanic-Serving Institutions: A national study. Journal of Hispanic Higher Education, 10, 18–40. https://doi.org/10.1177/1538192710391801

32.

Pew Research Center. (2020). U.S. Hispanic population surpassed 60 million in 2019, but growth has slowed. Author. https://www.pewresearch.org/fact-tank/2020/07/07/u-s-hispanic-population-surpassed-60-million-in-2019-but-growth-has-slowed/

33.

Postsecondary National Policy Institute. (2020). Latino students in higher education. Author. https://pnpi.org/latino-students/

34.

Sáenz

V. B.

de las Mercédez

Rodriguez

S. G.

Garcia-Louis

(2017). Latino men and their fathers: Exploring how Community Cultural Wealth influences their community college success. Association of Mexican American Educators Journal, 11, 89–110. https://doi.org/10.24974/amae.11.2.351

35.

Sáenz

V. B.

Ponjuán

Figueroa

(Eds.). (2016). Ensuring the success of Latino males in higher education: A national imperative. Stylus Publishing.

36.

Santiago

D. A.

Taylor

Calderón Galdeano

(2016). From capacity to success: HSIs, Title V, and Latino students. Excelencia in Education. https://www.edexcelencia.org/research/publications/capacity-success-hsis-title-v-and-latino-students

37.

Stephan

J. L.

Davis

Lindsay

Miller

(2015). Who will succeed and who will struggle? Predicting early college success with Indiana’s Student Information System. http://ies.ed.gov/ncee/edlabs

38.

Stewart

Lim

D. H.

Kim

(2015). Factors influencing college persistence for first-time students. Journal of Developmental Education, 38(3), 12–20.

39.

Tinto

(1975). Dropout from higher education: A theoretical synthesis of recent research. Review of Educational Research, 45, 89–125. https://doi.org/10.3102/00346543045001089

40.

Tinto

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

41.

Tinto

(2012). Completing college: Rethinking institutional action. University of Chicago Press.

42.

Willcoxson

Wynder

(2010). The relationship between choice of major and career, experience of university and attrition. Australian Journal of Education, 54, 175–189. https://doi.org/10.1177/000494411005400205

43.

Winterer

E. R.

Froyd

J. E.

Borrego

Martin

J. P.

Foster

(2020). Factors influencing the academic success of Latinx students matriculating at 2-year and transferring to 4-year US institutions – Implications for STEM majors: A systematic review of the literature. International Journal of STEM Education, 7, 34. https://doi.org/10.1186/s40594-020-00215-6

44.

Witkow

M. R.

Huynh

Fuligni

A. J.

(2015). Understanding differences in college persistence: A longitudinal examination of financial circumstances, family obligations, and discrimination in an ethnically diverse sample. Applied Developmental Science, 19(1), 4–18. https://doi.org/10.1080/10888691.2014.946030

45.

Wood

J. L.

Nevarez

Hilton

A. A.

(2012). Determinants of transfer among community college students. Journal of Applied Research in the Community College, 19(2), 63–68.