Remedial Education and Student Attrition.

Data from entering student surveys for the fall 1993, 1994, and 1995 freshmen cohorts at Utah Valley State College and 1998 outcomes data from the college's student information system were analyzed to determine student attrition rates and their relationship to variables defined in the literature about retention models, at-risk populations, and evaluating retention programs. Based on logistic regression analysis, the author discusses demographic, goal commitment, academic, and financial support variables as predictors of student retention.

Community colleges serve ninny students who are underprepared for their academic studies. According to the National Center for Education Statistics (NCES), approximately 41% of community college freshmen needed remedial education in 1995 (U.S. Department of Education, 1996). This lack of preparation creates greater challenges for students and the faculty and student personnel at community colleges who are trying to help them have a successful college experience. In many instances, community colleges provide a second chance for underprepared students to pursue higher education. The community college mission to assist these students, however, results in lower retention rates on campus. The present study examines the influence of student need for remediation on retention rates while controlling for other relevant variables.

Literature Review

Extensive research has been conducted on retention in higher education using different methods and approaches; most of the research, however, has been conducted at four-year colleges and universities rather than at community colleges. The following literature review provides an overview of these retention studies, which often have focused on particular areas of research such as the effect of student expectations, career decision-making, student financial aid, or the institutional environment. Some researchers have concentrated their efforts on special student populations such as older adults or minorities. Institutional researchers have also evaluated several factors simultaneously in efforts to predict student retention or to examine the relationships among several variables using conceptual models of student attrition. Finally, researchers have assessed whether specific intervention strategies reduce drop-out rates among college students. The literature in each of these areas is briefly discussed in the following sections.

General Models of Student Retention

Several theoretical models have been developed to explain the influences that affect a student's decision to complete their college studies. In Tinto's model, retention is influenced by a student's pre-entry attributes, goals and commitments, and academic and social integration (Tinto, 1975, 1993). Bean and Metzner (1985) developed a model conceptualizing student persistence as dependent on a student's background, academic variables, environmental variables such as employment and finances, and social integration. Cabrera and others (1992) integrated the Tinto model and the Bean and Metzner model, finding that both provided unique insights but also measured similar constructs.

Several researchers have used the Bean and Metzner or Tinto models to assess the impact of various factors on student retention at community colleges (Bets & Smith, 1991; Clagett, 1996; Feldman, 1993; Voorhees, 1993; Windham, 1995). Conflicting findings exist among many of these studies as to whether gender, student goals, need for remedial education, student grade point averages, contact with faculty, or hours studied can be related to student persistence. These studies have consistently revealed, however, that older students, part-time students, minority students, and working adults have higher drop-out rates.

These studies examined student retention over short time periods of one year or two semesters. Using a short time frame does not account for students who temporarily stop out and still may have a successful college experience later (Bonham & Luckie 1993; Grosset, 1993). In addition, student transfer to other institutions was not considered in several studies. If successful students are included in the group that is not retained, the results of retention studies may be confounded and result in conflicting findings.

The authors who accounted for student transfer found that the need for remedial education at community colleges significantly increased a student's risk of dropping out of college (Clagett, 1996; Vorhees, 1993). In contrast, researchers who did not account for student transfer found no relationship between remedial education and student persistence (Feldman, 1993).

Pascarella, Smart, and Ethington (1993) studied the degree persistence of 825 students from 85 different community colleges over a nine-year period. Academic and social integration were significant predictors of persistence for males and females. Socioeconomic status was an important factor for females, and institutional commitment or satisfaction was significant for males. Less than 26% of the variation in degree attainment was explained in the study.

Studies that use a theoretical framework and evaluate the relationship among several variables simultaneously offer more insight than those that only focus on one area because findings that appear significant may be spurious when considering other relevant factors. The efforts to confirm theoretical models of student attrition or predict student retention are incomplete, and much of the variation in retention rates among students remains unexplained.

Specialized Models

Along with using general models of student retention, institutional researchers have focused on specific areas of interest. The advantage of such studies is that they may provide new insights and suggest other factors that may influence attrition rates. For example, some researchers have examined student expectations and how they relate to student attrition (Bank, Biddle, & Slavings, 1992; Braxton, Vesper, & Hossler, 1995). These studies found that when the college expectations of students were met, the students' persistence or intent to persist increased; research in this area has explained a small percentage of the variance in student retention.

Another example is research on the relationship between career decision-making and student persistence. There is some evidence that a student's declaration of a college major or career relates to retention. Lewallen (1993) concluded that there was no relation between being initially undecided and retention when studying institutions of varied Carnegie classifications. Foote (1980), however, found a significant relationship when looking at a student's declared major during their experience at a university. Although Foote showed that students with undeclared majors had higher attrition rates, the relative importance of this variable in relation to other factors was not assessed in the study. Additional support is provided by Peterson (1993), who found that a student's career decision-making self-efficacy was related to academic and social integration at a state university.

Institutional environment or organizational characteristics have also been considered by researchers. Studies conducted in this area indicate that student feelings of alienation may be greater in large universities than in smaller educational institutions. Tomlinson-Clarke and Clarke (1996) found that men experienced more alienation and expressed more uncertainty than women in their decision to continue their studies. Students who lived on campus were shown to have a greater sense of community and higher retention rates (Lounsbury & DeNeui, 1995; Thompson, Smairatedu, & Rafter, 1993). Berger and Braxton (1998) stated that institutional communication, fairness in policy and decision-making, and participation were positively related to social integration and had significant indirect effects on attrition rates. Others have asserted that the quality of a student's experience in the classroom or the classroom environment was central to student retention (Ritschel 1995; Tinto 1997).

Billions of dollars in aid are awarded to students each year to "provide needy students with access to higher education, reasonable choices among alternative institutions, and financial ability to persist until graduation" (Wilcox, 1991, p. 48). Wilcox describes several research studies affirming the positive effect of financial aid on student persistence: Various combinations of aid have been shown to be effective with the exception of students who only received loans. Moline (1987) used multiple regression and path analysis to evaluate the effect of student financial aid at a liberal arts college when controlling for student characteristics, home proximity, adjusted income, and a student's grade point average. A student's GPA and high school rank had the largest effects. Nevertheless, "financial aid variables ... were not statistically significant in explaining persistence as measured by credits completed" when controlling for the other variables (Moline, 1987, p. 141). Just 35% of the variance in student persistence at the university was explained in the study.

Finally, researchers have studied inter-institutional differences in the retention rate at two-year colleges (Wyman, 1997, p. 32). The major factors affecting retention rates in Wyman's study were regional employment per capita and spending on instruction and academic support per student as a ratio of regional mean income. Institutions enrolling more full-time students also had lower attrition rates. Wyman likewise provided an informative literature review that discussed outcomes assessment and the importance of increasing retention rates on campus.

Although differences in retention rates among colleges and universities provide unique insights, educators also need to understand the needs of students enrolled on their campuses to improve retention rates. Higher education institutions have limited control over employment rates and funding levels in their states. In addition, community colleges accept a large number of part-time students and older working adults and operate under an open-admissions policy; their administrators and faculty have limited control over these factors. Given these conditions, institutional leaders continue attempts to discover ways to increase retention rates. Interventions often are based upon studies showing a need to address specific areas that affect individual student retention.

Studies on at-Risk Student Populations

Several studies have been conducted on special student populations that often have high drop-out rates. For example, more older working adults attend college than in the past. They may be returning to prepare for a second career or to update their knowledge and skills for the changing work environment. These students are older than the traditional 18- to 24-year-old college student who is preparing to enter the work force. Older adults are most often part-time students, and the large majority work while attending college. Researchers have found that persistence rates were lower for older adults at four-year institutions who worked more hours and only attended part-time (Naretto, 1995). They also have noted that a supportive social environment relates positively to the retention of older adults (Ashar & Skenes, 1993; Naretto, 1995). Richardson (1994) asserted that there are misconceptions that older adults generally lack academic skills and abilities. He reviewed a series of studies on older adults to support his argument.

Minority student groups have also been the subject of retention research because of their high attrition rates. For example, Nora and Cabrera (1996) found that minorities at a doctoral granting institution were somewhat less prepared for their college studies than White students and often believed that their campus environment was discriminatory. Based on their research, the authors reported that cumulative GPA was directly related to college persistence. Moreover, parental support, social integration, perceptions of discrimination, and academic and intellectual development were indirectly related to student retention.

Smedley, Hector, & Harrell (1993) used a series of instruments to assess sources of stress for minority students at a large university. They reported that academic stress and stress associated with minority status increased the risk of attrition for minority students. Minority students experienced stress related to academic achievement, which was correlated with lower grades. When matching African American and White students with similar academic achievement, Augelli & Hershberger (1993) found that African American students at the university had greater concerns with finances, had lower satisfaction with the institution, and frequently experienced verbal harassment. About 10% were threatened, 3% were assaulted, and 2% were spat upon. In a study of Hispanic students at several community colleges, Soils (1995) showed that satisfaction with instruction and academic experiences were related to a student's commitment to attend. Family support and future job prestige increased a student's motivation to persist.

Using the NCES High School and Beyond 1972 data, Adelman analyzed the transcripts of a national sample of college students to determine the effect of student need for remedial education on bachelor's degree attainment (Adelman, 1996). Students who completed a semester of credit were included in the analysis. Fifty-five percent of students who were college-ready earned a bachelor's degree, 45% of students who completed one remedial course earned a degree, but "only 24% of those who took three or more remedial courses had earned a bachelor's" (Adelman, 1996, p. A56). Students who required remedial reading were at high risk of needing remedial education in all three areas: remedial reading, English, and mathematics. Remediation rates were substantially higher among minorities, particularly for African Americans.

Evaluation of Student Retention Programs

In response to research on student attrition, colleges and universities have developed several intervention programs to increase their retention rates. Institutional researchers often use inventories to assess student needs when they enter the institution (Himelstein, 1992; Picklesimer & Miller, 1998; Ryland, Riordan, & Brack, 1994). If students are identified as high risk on the inventories, they are given additional assistance. Some researchers have supported the use of freshmen first-year programs such as seminars to help students learn study skills and understand college expectations and orientation sessions to link them with student support services (Fidler & Godwin, 1994; Glass & Garrett, 1995; Kluepfel, 1994; Singleton, Garvey, & Phillips, 1998; Wolfe, 1993).

The academic performance of students was improved at one university by identifying high risk courses that students frequently failed at the university (Kluepfel, Parelius, & Roberts, 1994). Each department developed gateway courses to prepare students for these courses. At another university, students on academic suspension took a study skills course and received additional advising, but the intervention did not improve subsequent success (Schultz, Dickman, Campbell, & Snow, 1992). Nevertheless, other researchers have demonstrated that study skills programs can be effective (Polansky, Horan, & Hanish 1993).

Several researchers have studied the efficacy of remedial education programs by evaluating course completion rates, progress from remedial to college-level work, subsequent performance in college-level courses, and comparing the persistence rates of remedial course completers with those who fail to complete remedial course work. Weissman, Bulakowski, and Jumisko (1997) provide an excellent review of several studies demonstrating the effectiveness of remedial education programs and offer important recommendations for structuring the education of underprepared students at community colleges. Simmons (1994) also provided data on the first-term performance of remedial education students involved in a summer transition program at a university. The study results supported the efficacy of the program.

Researchers also have focused on assessing and improving academic advising (Kern & Engels, 1996). At one university, students requested access to advising in the evenings, desired more information on career and job pursuits, needed more time with advisors, and wanted more assistance in learning study skills.

Colleges and universities sometimes establish a center or division for undeclared freshmen (Jones & Schultz, 1992; Toder & Hartsough, 1993). Faculty volunteers are used in these programs to provide advising to undeclared students. Toder and Hartsough used a control group design to demonstrate the effectiveness of the center at a state university.

Other programs used to increase retention include student orientation programs, support counseling groups, mentoring programs, honors programs, and service learning (Altizer & Patterson, 1994; Brown, 1994; Capello, 1994; Kish & Rita, 1993; Perkins, Paradowski, & Hirchert, 1994; Sax & Astin, 1997; Rao, 1998). Roueche and Roueche (1994) recommend mandatory orientation programs and basic skills assessment, completion of remedial work before regular college level courses, abolishing late registration, establishing mentoring and peer support programs, reducing academic loads for working students, providing more financial aid, and conducting more program evaluations.

As demonstrated by the sometimes conflicting findings in the literature, studying student retention remains difficult and complex. Many possible factors can directly and indirectly influence a student's decision to persist. For this reason, both two- and four-year institutions use several interventions in their attempts to reduce attrition. An effective student retention program requires a campus-wide effort, and can involve several different student sub-populations. Specific influences may be different for each group.

The study described in this report was designed to investigate many different variables defined in the existing literature to determine their relationship to retention rates at one community college. The overall goal was to add insights and descriptive detail about the relationship between remedial education and student persistence rates at community colleges.

Method

The study was conducted at Utah Valley State College (UVSC), an urban community college that enrolled 18,174 students during the fall of 1998. Historical data maintained on the college indicate that over half of the students drop-out of the college and fail to earn a degree or transfer, and about half the entering freshmen require remedial education. The average age of students at the college is 22. About 54% are males, and 46% females, with 93% White. With the exception of having fewer minority students (7% compared with 24% at two-year colleges nationally), this population is similar to the national population of students at community colleges where approximately 62% of those who enroll fail to earn any degree or certificate and drop-out of college (NCES, 1996, p. 3, 15).

The fall 1993, 1994, and 1995 freshmen cohorts were tracked to determine how many students graduated, transferred, were still enrolled, or dropped out of the institution by fall 1998. Entering freshmen were identified by obtaining a student's first-term taking a college course on the campus. Students transferring in with credit from other institutions were eliminated from the cohorts. Information from the student information system, a new student survey, records from the Utah State Board of Regents (USBR), and records from Brigham Young University (BYU) were used to identify students transferring from other institutions. Students classified as re-admitted in the student information system were also excluded from the cohorts. Concurrent or dual enrollment students were included only if they attended a regular college course on campus.

Students who received a degree or certificate were classified as graduates. Records from the USBR and BYU were used to identify students who transferred to another college or university in the state before earning a degree or certificate at the community college. Students may have transferred out of state, but records to determine this were not available for the study. About 15% of entering freshmen transfer to other public universities or to Brigham Young University in the state after five years. The percentage of students transferring to out-of-state universities would be much smaller because of high nonresident tuition rates. Students were categorized as still enrolled at the college if they were taking courses during fall 1998. All other students were considered drop-outs or noncontinuing students.

Logistic regression was used to gain an understanding of which variables had the strongest direct relationship with student attrition. Because of indirect relationships and inter-correlation among the variables, it is difficult to assess the importance of individual factors. Nevertheless, logistic regression is an appropriate method used to study student retention and provides a rough indication of which factors are likely to have the largest direct relationships (Dey & Astin, 1993). The variables in the equation were entered using forced entry and obtained from the student information system and new student survey. Definitions of the variables are provided in Table 1.

Table 1

Definition of Variables

Demographic Variables

Older                  More than age 24 at start of first regular
                       term at the college (0=No, 1=Yes)
Minority               Black, Hispanic, Asian, American Indian, and
                       Alaskan native students (0=No, 1=Yes)
Disabled               Identified as disabled (0=No, 1=Yes)
Single parent          Identified as single parent (0=No, 1=Yes)
First generation(*)    Mother or father completed a four-year degree
                       or higher (0=No, 1=Yes)
Concurrent             Took concurrent enrollment courses while in
                       high school (0=No, 1=Yes)
Married                Identified as married (0=No, 1=Yes)
Children(*)            Identified as having children (0=No, 1=Yes)
Attempted hours(*)     Number of attempted credit hours during first
                       regular term at the college

Goal Commitment Variables

Bachelor's degree(*)   5-point Likert scale where 1=Very likely and
                       5=Very unlikely to pursue a four-year degree
Planned semesters(*)   Number semesters planned to attend
                       (range of 1 to 8+)

Academic Variables

Remedial 1 area        Students who tested into one remedial area
                       based on Compass and DRP test scores
                       (0=-No, 1=Yes)
Remedial 2 areas       Students who tested into two remedial areas
                       (0=No, 1=Yes)
Remedial 3 areas       Students who tested into three remedial areas
                       (0=No, 1=Yes)
GPA B range            Earned a first-term GPA ranging from 2.7
                       to 3.69
GPA C range            Earned a first-term GPA ranging from 1.7
                       to 2.69
GPA below C            Earned a first-term GPA below 1.7

Financial Support Variables

Received aid           Received short-term or long-term loans,
                       grants, or scholarships (0=-No, 1=Yes)
Lowest income(*)       Annual family income < $12,000
Second income(*)       Annual family income $12,000 - $24,999
Third income(*)        Annual family income $25,000 - $29,999
Fourth income(*)       Annual family income > $29,999
Work part-time(*)      Working 1 - 30 hours per week while
                       attending college
Work full-time(*)      Working 31 or more hours per week while
                       attending college
Live with parents(*)   Living with parents (0=No, 1=Yes)

(*) Data obtained from the new student survey. All other data were obtained from the college's student information system.

Overall Retention and Remediation Rates

When examining data on the fall 1993 to 1995 cohorts through fall 1998, student attrition rates ranged from 54 to 64%. Depending on the cohort, the college generally lost 30 to 35% of its students from fall to spring and nearly 60% of its students by the following fall. Some of these students were stop-outs and returned to continue their studies.

There are no admission requirements to attend the college because it is an open admissions institution. The college does, however, operate a testing program to identify students who need remedial education. The college offers several remedial math, English, and reading courses to bring students up to a college level of performance.

As the number of remedial areas increased for students at the college, their drop-out rates consistently increased (see Table 2). For example, 64 to 72% of students who required remedial education in three areas eventually dropped out of the college. In other words, high remediation rates had a negative relationship with student retention.

Table 2

Student Drop-Out Rates by Cohort and Remedial Category as of Fall 1998

                          Fall Cohort Drop-Out Rates

              1993 (5 Years)   1994 (4 Years)   1995 (3 Years)
Category        N = 2,140         N=2,659          N=2,888

Nonremedial         51               59               60
One area            55               60               63
Two areas           58               67               70
Three areas         64               68               72

Note. Values represent percentages of students in each category.

Many students at the college needed remedial education in several areas. About 21% of students required remedial education in two areas, and 11% required remedial education in three areas. About 44% needed remedial math courses, 34% needed remedial English, and 12% needed remedial reading courses. Students requiring remedial reading courses generally needed remediation in all three areas. These results confirmed the findings of Adelman's High School and Beyond study (1996).

Remedial education was related to a student's GPA during their first-term at the college. The average first-term GPA for non-remedial students was 2.8. Remedial education students had an average first-term GPA of 2.54, with most students earning C grades in courses. Those needing two areas of remedial education had a first-term GPA of 2.47, and those needing remedial courses in all three areas had an average first-term GPA of 2.30. In other words, a lack of preparation for college substantially reduces a student's chances of college success by decreasing their ability to perform academically. A large remedial population at the college creates a considerable challenge for the institution.

Some student groups with high drop-out rates had high remedial placement rates. For example, disabled students had the highest remedial placement rates of all the student groups compared. About 72% required remedial education, 28% needed remediation in two areas, and 22% needed remediation in three areas. About 64% required remedial math, 53% remedial English, and 27% remedial reading.

Minorities also had high remedial placement rates. Sixty-two percent tested into remedial education, with 57% requiring remedial math, 51% remedial English, and 26% remedial reading. More than twice as many minorities as Whites needed remedial education in three areas.

Regression Model

The 1993 to 1995 cohorts were combined in the regression analysis to identify factors that related significantly to the retention of remedial education students over these years (see Table 3). A second regression was also run for the data available on all students to see if findings were similar. Regression analysis is influenced by the variables that are included in the model; nevertheless, the direction of the relationships was consistent for all students and remedial education students.

Table 3

Estimates of Retention Equations for Remedial and All Students

                        Remedial Students
                           (N = 2,517)

Variables              Coefficients     R

Demographic
Older                    .2017        .0000
Minority                 .4428(*)     .0323
Disabled                 .1216        .0000
Single Parent            .4541        .0065
First Generation        -.1874(*)    -.0254
Concurrent enrollment   -.6076(**)   -.0731
Married                  .1738        .0000
Children                -.0348        .0000
Attempted hours         -.0393(**)   -.0612

Goal Commitment
Bachelor's degree        .0128        .0000
Planned semesters        .0753(**)    .0517

Academic
Remedial one area
Remedial two areas       .0833        .0000
Remedial three areas     .0643        .0000
GPA B range             -.0095        .0000
GPA C range              .3930(**)    .0436
GPA lower than C        1.2585(**)    .1247

Financial Support
Received aid            -.4780(**)   -.0838
Lowest income group      .4264(*)     .0324
Second income group     -.0320        .0000
Third income group       .3998(**)    .0401
Work part-time           .1354        .0000
Work full-time           .4875(**)    .0407
Live with Parents       -.2464(**)   -.0375
Constant                 .5004

                          All Students
                          (N = 7,683)

Variables              Coefficients    R

Demographic
Older                    .8413(**)    .0684
Minority                 .5215(**)    .0407
Disabled                 .1175        .0000
Single Parent
First Generation
Concurrent enrollment   -.5397(**)   -.0755
Married                  .2634(**)    .0232
Children
Attempted hours         -.0411(**)   -.0708

Goal Commitment
Bachelor's degree
Planned semesters

Academic
Remedial one area        .0835        .0000
Remedial two areas       .2482(**)    .0335
Remedial three areas     .2551(**)    .0234
GPA B range             -.0113        .0000
GPA C range              .4425(**)    .0557
GPA lower than C        1.2188(**)    .1224

Financial Support
Received aid            -.5358(**)   -.1029
Lowest income group
Second income group
Third income group
Work part-time
Work full-time
Live with Parents
Constant                 .7394(**)

(*) p < .05 (**) p < .01

Additional information was available on remedial education students because they completed a new student survey when taking placement tests at the college. Only students with low ACT tests are required to complete the testing. The new student survey gathers information that has been supported by prior studies as being important to student retention. The variables were included in the model based upon theoretical arguments that they related to student retention. This contrasts with other methods of entering a large number of variables that might have a relationship, many of which have no significant relationship with the dependent variable and may result in spurious correlations.

The regression equation for the remedial population included 2,517 students, and the equation for the overall population included 7,683 students. Students who dropped out as of 1998 were coded with a 1, and students who graduated, transferred, or were still enrolled were coded with a 0.

Several variables used in file model were entered as categorical variables. In these instances, a comparison group was created. For example, separate dummy variables (coded as 0 or 1) were created for students who needed remedial education in one area, two areas, and three areas. These variables were entered into the regression separately and compared with students not requiring any remedial education.

This same method was used for student first-term GPAs, income groups, and working students. Students with GPAs in the B range, C range, and below a C were compared with students earning A grades. Students from the lowest, second, and third income groups were compared with students with the highest incomes. Students working full-time and part-time were compared with students who did not work.

The regression coefficients and partial correlations (R) are presented in Table 3. A positive coefficient indicates an increase in attrition, whereas a negative coefficient implies that the factor may contribute to student success. The partial correlations estimate the strength of the relationship between the independent and dependent variable, controlling for the effect of the other independent variables in the model. Partial correlations with larger absolute values suggest that the variables have a stronger direct relationship with the dependent variable.

Several findings were consistent for all students and remedial education students. The variable that had the largest partial correlation in both equations involved a student's GPA during their first-term at the college. Because the distribution for GPA was bimodal, grade information was entered in the equation as categorical data. The retention rates for students with GPAs in the B range were not significantly different from A students. The positive coefficients, however, indicate that students with GPAs in the C range and below during their first-term were at a significantly higher risk of dropping out of college than students with higher GPAs, when controlling for all the other factors in the model. In other words, academic performance appears to be the most important factor.

As shown in the descriptive statistics presented above, remedial education was associated with low first-term GPAs. When controlling for other variables in the overall model, students who needed remedial education in two and three areas had significantly higher drop-out rates than students who did not need remedial education. The level of remediation, however, was not significant in the remedial student model. By examining the retention of only underprepared students, researchers may not be able to detect the relationship between remedial education and retention.

The dummy variable measuring receipt of financial aid had the second largest partial correlation in both models. The Wald statistic on the negative coefficient was significant and indicated that the receipt of financial aid reduced dropout rates when controlling for the other factors.

Other variables related to the financial support of students indicated that this area was important in predicting student retention. For remedial education students, living at home had a significant and negative relationship with student drop-out rates. There is often concern with students working full-time while attending college because they may have less of a commitment to their academic studies. Students from lower socioeconomic groups may need to work more hours than students from higher income groups to pay for college. In the regression, students that worked full-time and were from lower income groups had significantly higher drop-out rates. Working part-time did not appear to have a significant impact on student attrition.

If students work full-time, they will generally take fewer courses because of work demands on their time. In this study, students who took a larger number of credit hours at the college during their first-term had higher retention rates. Full-time students may have a greater commitment or more financial support for pursuing their studies.

Student demographics were also significant predictors of student retention rates. When controlling for first-term GPA, remedial placement rates and other variables, minority status still significantly increased a student's chances of dropping out of college. Depending on the cohort 73 to 77% of minority students on campus dropped-out, failing to complete their studies. Although these students had higher remedial education needs, there may be other factors that influence their retention at the college.

Being older than 24 when starting college and married was significant in the overall model but not for remedial education students. This may be the case because the equation for remedial education students included the hours that students worked while attending college. For remedial education students, single parents had high drop-out rates ranging from 66 to 87% depending on the cohort, but single parenthood was not significant in the regression. Other factors in the model may have accounted for their higher drop-out rates.

Students with a parent who earned a bachelor's degree were more likely to persist, whereas first-generation students had higher attrition rates. The number of semesters that a student planned to attend the college was also a significant predictor of retention. It was interesting to note that students who planned to attend more semesters at the college had lower retention rates.

The impact of dual enrollment was included in the study to assess its possible effect on student retention. Students who completed concurrent enrollment courses in high school had higher retention rates.

The remedial student equation and overall equation correctly classified 66% and 65% of the students respectively. When including all the variables in the remedial student equation, 85% of students who dropped out were classified correctly; in contrast, just 37% of those who persisted were classified correctly. The model had difficulty identifying students who persisted at the college.

The research supported the findings of most studies showing that academic performance, minority status, work, and other outside commitments significantly relate to student retention. The level of remedial education required by students also related to their academic performance and persistence. In addition, the effects of living at home and early exposure to college through concurrent enrollment were considered--variables that have not been included in other studies at community colleges that proved to be significant.

Implications For Research and Practice Predicting student retention is difficult because so many factors influence a student's college success. These include students' characteristics, their commitment to earning a degree, academic variables, student financial support, and other factors. Nevertheless, the study found that a student's first-term academic performance had the strongest relationship with student retention of all the variables considered. The high remedial population of the college significantly increased drop-out rates, and influenced the overall student GPA. The second most influential factor was student receipt of financial aid.

This study supports the need for interventions that focus on the academic needs of students. These include remedial education courses, early warning systems, transition programs, study skills courses, and the use of gateway courses. Academic advisors can assist students in making appropriate decisions about the remedial courses they need or help students understand the effect of outside work commitments on the achievement of their educational goals. Students, may need encouragement and assistance in obtaining an adequate amount of financial aid. The quality of instruction and interaction between students and faculty in the classroom is vitally important. The effectiveness of support programs, learning outcomes, and the quality of instruction at community colleges should be examined.

In addition to retention programs on college campuses, a student's high school studies affect their academic preparation for college. Students who have to complete several remedial courses may become discouraged, experience difficulty in academic subjects, and drop out of college. Community colleges and high school districts should develop closer partnerships to improve the rigor and quality of high school programs. High schools should adopt higher standards to increase the preparation of students for their college experience (Hoyt, 1999a, 1999b).

Another way to improve student retention is to strengthen the financial aid programs on college campuses. This is critical for low income students who have limited family support. College student personnel should also recognize the benefits that students receive by living at home and not necessarily promote residential living on campus or elsewhere. Living at home may be less expensive and thus less stressful for students and their families, resulting in higher persistence rates.

Retention rates may be greater for students planning to attend fewer semesters at the community college for several reasons. Students who planned to complete short-term certificate programs may have been more successful. In other words, underprepared students may be able to achieve short-term goals easier than long-term goals. Advisors may need to encourage these students to complete short-term certificate programs and build upon them as they earn higher levels of education and training. Students planning to attend more semesters may also be studying part-time and planning to take longer to earn their degree. Work and family commitments may reduce a student's ability to sustain educational commitments over a long period of time. Prospective transfer students, who planned to only attend a few terms, may have been less likely to drop-out of the community college. This underscores the importance of considering transfer rates and evaluating the subsequent success of students at receiving institutions.

Taking concurrent enrollment courses may prepare students for college or demonstrate a stronger commitment to earning a degree. These programs introduce students to college-level curriculum, allow them to expand their educational experiences, and give them a head start in college. This creates an early investment in higher education, which may reduce time to degree completion and increase individual commitment through college.

First-generation college students may be less likely to persist due to a lack of support from home. Parents who have not earned a four-year degree may not fully appreciate the value of higher education nor expect their children to finish their studies. These parents may lack the economic means to assist their children financially. Parents without college experience may be less informed about the process and less able to guide and support their children through the college experience. This supports the need for programs to assist these students.

Minorities represented a small proportion of the student population at the college where this research took place; nevertheless, they were at much greater risk of requiring remedial education in several areas and dropping out of college. As the minority population in this country increases, the need to address its preparation for college will become even more important for the educational community. The college's responsibility for promoting a positive campus environment free of discrimination will continue. The study findings that show very high drop-out rates among these students supports the need for minority student centers and support programs.

Although the need for remedial education has a significant relationship with student retention, many other factors influence student retention. Because of this, several programs and interventions are needed to increase student retention. Improving student retention does not have a simple easy answer, and it requires a campus-wide effort.

References

Adelman, C. (1996, October 4). The truth about remedial work. The Chronicle of Higher Education, p. A56.

Altizer, A.A., & Patterson, C.J. (1994). Outdoor orientation: Sailing beyond the formal domain. Journal of College Student Development, 35, 135.

Ashar, H., & Skenes, R. (1993). Can Tinto's student departure model be applied to non-traditional students? Adult Education Quarterly, 43(2), 90-97.

Augelli, A.R., & Hershberger, S.L. (1993). African American undergraduates on a predominantly White campus: Academic factors, social networks, and campus climate. Journal of Negro Education, 62(1), 67-81.

Bank, B.J., Biddle, B.J., & Slavings, R.L. (1992). What do students want?: Expectations and undergraduate persistence. The Sociological Quarterly, 33(3), 321-329.

Bean, J.P., & Metzner, B.S. (1985). A conceptual model of non-traditional undergraduate student attrition. Review of Educational Research, 55(4), 485-540.

Berger, J.B., & Braxton, J.M. (1998). Revising Tinto's interactionalist theory of student departure through theory elaboration: Examining the role of organizational attributes in the persistence process. Research in Higher Education, 39(2), 103-119.

Bers, T.H., & Smith, K.E. (1991). Persistence of community college students: The influence of student intent and academic and social integration. Research in Higher Education, 32(5), 539-556.

Bonham, L., & Luckie, J. A. (1993). Community college retention: Differentiating among stopouts, dropouts, and optouts. Community College Journal of Research and Practice, 17, 543-554.

Braxton, J.M., Vesper, N., & Hossler, D. (1995). Expectations for college and student persistence. Research in Higher Education, 36(5), 595-611.

Brown, S. (1994). Images of a model to promote retention of Black female students on predominantly White campuses. Journal of College Student Development, 35, 150-151.

Cabrera, A.F., Castaneda, M.B., Nora, A., & Hengstler, D. (1992). The convergence between two theories of college persistence. Journal of Higher Education, 63(2): 143-164.

Capello, D.C. (1994). Beyond financial aid: Counseling Latina students. Journal of Multicultural Counseling and Development, 22, 28-35.

Claggett, C.A. (1996). Correlates of success in the community college: Using research to inform campus retention efforts. Journal of Applied Research in the Community College, 4(1): 49-68.

Dey, E.L., & Astin, A.W. (1993, December). Statistical alternatives for studying college student retention: A comparative analysis of logit, probit, and linear regression. Research in Higher Education, 34, 569-581.

Feldman, M.J. (1993). Factors associated with one-year retention in a community college. Research in Higher Education, 34(4), 503-511.

Fidler, P.P., & Godwin, M.A. (1994). Retaining African-American students through the freshmen seminar. Journal of Developmental Education, 17(3), 34-38.

Foote, B. (1980). Determined and undetermined-major students: How different are they? Journal of College Student Personnel, 21(1), 29-31.

Glass, J.C. Jr., & Garrett, M.S. (1995). Student participation in a college orientation course, retention, and grade point average. Community College Journal of Research and Practice, 19(2), 117-132.

Grosset, J.M. (1993). A profile of community college stop-outs. Community College Review, 20(4), 51-58.

Himelstein, H.C. (1992). Early identification of high risk students: Using noncognitive indicators. Journal of College Student Development, 33(1), 89-90.

Hoyt, J.E. (1999a, Spring). Level of math preparation in high school and its impact on remedial placement at an urban state college. College and University, 74(3), 37-43.

Hoyt, J.E. (1999b). Promoting academic standards: The link between remedial education in college and student preparation in high school. Orem, UT: Department of Institutional Research, Utah Valley State College.

Jones, E.G., & Schultz, M.B. (1992). An academic advising program for undeclared students. Journal of College Student Development, 33(1), 181-183.

Kern, C.W., & Engels, D.W. (1996). Developmental academic advising: A paradigm shift in a college of business administration. Journal of College Student Development 37(1), 95.

Kish, E.R., & Santa Rita, E. (1994). The minority honors program in energy-related curricula. College and University, 17(3), 80-85.

Kluepfel, G.A. (1994). Developing successful retention programs: An interview with Michael Hovland. Journal of Developmental Education, 17(3), 28-33.

Kluepfel, G.A., Parelius, R.J., & Roberts, G. (1994). Involving faculty in retention. Journal of Developmental Education, 17(3), 16-24.

Lewallen, W (1993, March). The impact of being undecided on college-student persistence. Journal of College Student Development, 34, 103-113.

Lounsbury, J.W., & DeNeui, D. (1995). Psychological sense of community on campus. The College Student Journal, 29(3), 270-277.

Moline, A.E. (1987). Financial aid and student persistence: An application of causal modeling. Research in Higher Education, 26(2), 130-147.

Naretto, J.A. (1995). Adult student retention: The influence of internal and external communities. NASPA Journal, 32(2), 90-100.

National Center for Education Statistics. (1996, May). Descriptive summary of 1989-90 beginning postsecondary students: Five years later. Washington, DC: Office of Educational Research and Improvement.

Nora, A., & 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.

Pascarella, E.T., Smart, J.C., & Ethington, C.A. (1993). Long-term persistence of two-year college students. Research in Higher Education, 24(1), 47-71.

Perkins, R.J., Paradowski, J.W., & Hirchert, S.E. (1994). The MEGA experience: A mentoring program that works. Journal of College Student Development, 35, 152.

Picklesimer, B.K., & Miller, T.K. (1998). Life-skills development inventory-college form: An assessment measure. Journal of College Student Development, 39(1), 100-110.

Peterson, S.L. (1993). Career decision-making self-efficacy and institutional integration of underprepared college students. Research in Higher Education, 34(6), 659-685.

Polansky, J., Horan, J. J., & Hanish, C. (1993). Experimental construct validity of the outcomes of study skills training and career counseling as treatments for the retention of at-risk students. Journal of Counseling and Development, 71, 489492.

Rao, M. (1998, January/February). Mentoring and peer relationships. Change, pp. 4648.

Richardson, J.T. (1994, October). Mature students in higher education: Academic performance and intellectual ability. Higher Education, 28(3), 373-386.

Ritschel, R.E. (1995). The classroom as community. Community College Journal, 65(4), 16-19.

Roueche, J.E., & Roueche, S.D. (1994). Responding to the challenge of the at-risk student. Community College Journal of Research and Practice, 18(1), 111.

Ryland, E.B., Riordan, R.J., & Brack, G. (1994, January). Selected characteristics of high-risk students and their enrollment persistence. Journal of College Student Development, 35, 54-58.

Sax, L.J., & Astin, A.W. (1997, Summer/Fall). The benefits of service: Evidence from undergraduates. Educational Record, pp. 24-33.

Schultz, R.A., Dickman, M.M., Campbell, N.J., & Snow, B.M. (1992). Assessing a short-term intervention to facilitate academic success. NASPA Journal, 30(1), 43-50.

Singleton, R.A. Jr., Garvey, R.H., & Phillips, G.A. (1998, May/June). Connecting the academic and social lives of students. Change, pp. 18-25.

Simmons, R. (1994). Pre-college programs: A contributing factor to university student attrition. Journal of Developmental Education, 17(3), 42-45.

Smedley, B.D., Hector F.M., & Harrell, S.P. (1993, July/August). Minority-status stresses and the college adjustment of ethnic minority freshmen. Journal of Higher Education, 64(4), 434-452.

Solis, E. Jr. (1995). Regression and path analysis models of hispanic community college students' intent to persist. Community College Review, 23(3), 3-15.

Thomlinson-Clarke, S., & Clarke, D. (1996). Institutional effects on alienation and student effort at three campus environments. Journal of College Student Development, 37(1), 60-67.

Thompson, J., Samiratedu, V., & Rafter, J. (1993, Fall). The effects of on-campus residence on first-time college students. NASPA Journal, 31(1), 4147.

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

Tinto, V. (1993). Leaving college. Chicago: The University of Chicago Press.

Tinto, V. (1997, November/December). Classrooms as communities. Journal of Higher Education, 68(6), 599-623.

Toder, F., & Hartsough, C.S. (1993, March). A collaborative approach to retain first-year undeclared students. Journal of College Student Development, 34, 159.

U. S. Department of Education, National Center for Education Statistics. (1996). Remedial education at higher education institutions in fall 1995. Washington DC: U.S. Government Printing Office.

Voorhees, R. A. (1993). Toward building models of community college persistence: A logit analysis. Research in Higher Education, 26(2), 115-129.

Weissman, J., Bulakowski, J., & Jumisko, M.K. (1997). Using research to evaluate developmental programs and policies. In Ignash, J.M. (Ed.), Implementing effective policies for remedial and developmental education (pp. 73-80). New Directions for Community Colleges, No. 100. San Francisco: Jossey Bass.

Wilcox, L. (1991). Evaluating the impact of financial aid on student recruitment and retention. In Hossler, D. (Ed.), Evaluating student recruitment and retention programs (pp. 47-60). New Directions for Institutional Research, No. 70. San Francisco: Jossey-Bass.

Windham, Patricia. (1995). The relative importance of selected factors to attrition at a public community college. Journal of Applied Research in the Community College, 3(1), 65-78.

Wolfe, J.S. (1993, Spring). Institutional integration, academic success, and persistence of first-year commuter and resident students. Research in Higher Education, 34, 321-326.

Wyman, F.J. (1997). A predictive model of retention rate at regional two-year colleges. Community College Review, 25(1): 29-45.

Jeff Hoyt serves as manager of institutional analysis at the University of Utah. ([e-mail])