Community College Reverse Transfer Students: A Field Survey of a Nontraditional Student Group.

Data from a 54-item survey directed at reverse transfer students in the University of Kentucky Community College System was analyzed based on responses from two subgroups: 734 completers (those who had completed a baccalaureate) and 148 noncompleters. Chi square and t tests indicated statistically significant differences between completers and noncompleters as to race and marital status, age, number of dependents, number of credit hours, and grade point average. The data also revealed differences between the two groups as to reasons for enrollment and current goals. The authors discuss the resulting profile of reverse transfers in a student services context.

Community college educators have been aware for more than two decades that the profile of community college students is undergoing steady and profound change. A facet of this change has been that students described as "nontraditional" constitute an increasing proportion of the student population. Nontraditional students (Cohen & Brawer, 1996) are individuals who do not conform to the profile of the traditional 18-year-old student who enrolls full-time at a community college, completes the freshman and sophomore years, and transfers to a four-year college to earn a baccalaureate degree.

Cohen and Brawer (1996) noted various changes that during the period 1970 to 1994 affected the, nontraditional student population: (a) the mean age for students increased from 27 in 1980 to more than 31 by 1993 as large numbers of adult learners returned to college to acquire and upgrade skills; (b) females, many of whom attend college part-time, did not equal males in enrollment until 1978 but outnumbered males (55% to 45%) by 1991; (c) minority enrollment increased from 20% in 1976 to 25% by 1991; and (d) part-time students, most of whom are members of one or more nontraditional groups, increased from 49% of the student population in 1970 to more than 65% of the population by 1992. The above patterns have remained the same through 1996, and it is likely that over 65% of the students enrolled in community colleges fall into at least one nontraditional student category (American Council on Education, 1998).

Another nontraditional student group that has been growing since 1970 is the group described in the community college literature as reverse transfer students. Kajstura and Keim (1992) operationally defined reverse transfers as "individuals who, prior to attending a two-year college, were last enrolled at a four-year institution" (p. 39). These researchers described reverse transfers further as belonging to one of two subgroups: "1) non-completers, who attended a four-year institution, but did not complete a degree before enrolling at a two-year college; and 2) graduates, who earned at least an undergraduate degree prior to enrolling at a two-year college" (Kajstura & Keim, 1992, p. 39).

Although the number of reverse transfers is not known with precision, previous research (Hogan, 1986; Mitchell & Grafton, 1985) indicates reverse transfers may comprise as much as 20% of community college enrollments. Even if only an average percentage from the literature (12%) is applied to the population of 5.2 million students enrolled in community colleges for credit (American Association of Community Colleges, 1997), the number of reverse transfers exceeds 600,000 students. Research about reverse transfers is needed because despite the emergence of a small body of reverse transfer literature, "reverse transfer students are a little-studied segment of the college student population [and] such students, who transfer from four-year colleges and universities to community colleges, represent a large uncharted population" (Swedler, 1983, p. 131).

Reverse transfer inquiry is warranted for three additional reasons. First, various studies suggest reverse transfer activity is increasing (Brimm & Achilles, 1976; Clark, 1960; Clark, 1982; Cohen, Palmer, & Zwemer, 1986). Second, the resulting changes in the profile of community college students affect existing programs and policies for student recruitment and retention, student services, and curriculum and instruction. Third, much previous research about reverse transfers has been limited to single institutions (Klepper, 1990; Ross, 1982), or to only a few community colleges within a state system (Lee, 1975; Renkiewicz, Hirsch, Drummond, & Mitchell, 1982). Analyses across all institutions within a state system are desirable because reverse transfer enrollments, as well as reverse transfer characteristics, vary by state (Hogan, 1986; Klepper, 1990; Kuznik, Macey, & Anderson, 1974; Mitchell & Grafton, 1985; Renkiewicz et al., 1982).

The investigation described in this report had two objectives. One objective was to develop a descriptive profile of completer and noncompleter reverse transfers to guide implementation of student services, programs, and advocacy for reverse transfers. The second objective was to develop reliable data that could be generalized statewide by selecting a sample from the population of all reverse transfers in Kentucky. Three research questions guided the investigation: (a) What are the characteristics of reverse transfers within the focal population? (b) What are the reasons for initial enrollment and current goals that motivate reverse transfers to attend community colleges? (c) How do completer and noncompleter reverse transfers differ with respect to their characteristics and with respect to their initial reasons for enrollment and current goals for attending a community college?

Methodology

The framework for this research was a specified student services model, the Services, Programs, Advocacy, Research (SPAR) Model developed by Jacoby and Girrell (1981). The SPAR Model was designed to address the needs of special student groups composed primarily of commuter students. The assumption underlying the SPAR Model is that "research regarding [student] characteristics and needs is the foundation upon which services, programs and advocacy efforts are developed" (Barr & Associates, 1993, p. 476). The specific component of the SPAR Model that guided this investigation was the research component, which emphasizes the field survey as the primary method for accessing information about special student groups. The field survey was implemented according to procedures recommended by Dillman (1978) and Fowler (1988).

The investigators extended previous reverse transfer research by implementing two advancements relative to sampling procedure. The first advancement was to implement a statewide sample; the second was to conduct a power analysis to ensure an adequate sample size. Reverse transfer researchers generally fail to report the method used to determine the sample size. Accordingly, the findings of some investigations, due to inadequate sample sizes, may have been flawed by a type II statistical error--that is, failing to find statistical significance when, in fact, significance exists (Cohen, 1988; Keppel, 1991).

Participants and Sampling Procedures

The participants (N= 882) in this study were reverse transfer students identified from the population of all reverse transfers enrolled for credit at the 14 community colleges within the University of Kentucky Community College System. The research sample included two subgroups: noncompleter reverse transfers (n=734) and completer (completed a bachelor's degree) reverse transfers (n=148). The data source used to identify the total population of reverse transfers was a list obtained from the central office of the University of Kentucky Community College System.

The research plan included two additional requirements: (a) a minimal survey response rate and (b) a minimal sample size to ensure sufficient statistical power. The response rate criterion was 60%, as recommended by Babbie (1992) and Dillman (1978). The sample size resulted from two steps: (a) a power analysis conducted according to procedures explicated by Gall, Borg and Gall (1986) and (b) use of an average for completer reverse transfers (12%) derived from previous research. The 12% factor was entered into the survey mailing computation described below to ensure the survey response would render an adequate sample size for the completer reverse transfer group.

The power analysis procedures (Gall, Borg, & Gall, 1986) yielded required sample sizes for both the completer reverse transfers (n= 100), and noncompleter reverse transfers (n=733). The power analysis specifications were: (a) power=.80, (b) alpha=.05, and (c) medium effect size (r=.06). Application of the estimated percentage for completer reverse transfers (12%), the planned sub-sample size for completer transfers (n=100), and the specified response rate (60%) yielded a required mailing of 1,389 questionnaires (100/.12/.60 = 1,389). The questionnaire was mailed to 1,389 potential respondents selected at random from a list of all reverse transfers in the focal population. This procedure rendered a randomized mailing but did not eliminate potential participant response bias. Descriptive statistics for the respondents appear in Table 1 (see pages 20).

Table 1 Descriptive Statistics for Study Participants (N = 882)

Variable                          M      SD    Range

Noncompleter Reverse Transfers
  Age                            29.3    8.9    19-67
  Gender                          (a)    (a)     (a)
  Race                            (b)    (b)     (b)
  Dependent children               .6     .9     0-4
  Hours enrolled                  9.0    4.3     0-19
  Hours completed                32.5   25.1     0-99
  GPA                             3.2     .6   1.00-4.00

Completer Reverse Transfers
  Age                            37.5   12.2    20-77
  Gender                          (a)    (a)     (a)
  Race                            (b)    (b)     (b)
  Dependent children               .8    1.4     0-5
  Hours enrolled                  5.4    3.7     0-15
  Hours completed                23.1   26.1     0-99
  GPA                             3.5     .5   1.00-4.00

Total
  Age                            30.7   10.1    19-77
  Gender                          (a)    (a)     (a)
  Race                            (b)    (b)     (b)
  Dependent children                     1.0     0-5
  Hours enrolled                  8.4    4.4     0-19
  Hours completed                30.1   25.6     0-99
  GPA                             3.2     .6   1.00-4.00

(a) The noncompleter reverse transfer distribution was 32.4% males and 67.6% females, and the completer reverse transfer distribution was 39.2% males and 60.8% females.

(b) The noncompleter reverse transfer distribution was 87.1% White and 12.9% minority, and the completer reverse transfer distribution was 93.2% White and 6.8% minority.

Instrumentation

The survey instrument was a questionnaire developed by the California Association of Community Colleges' Commission on Instruction and Research and Development. Renkiewicz et al. (1982) used the instrument first, and Klepper (1990) revised the instrument for a subsequent study. After pilot testing (explained below), a revised version of the instrument used by Klepper (1990) served as the instrument for this research. The instrument contained 54 items in four areas: (a) student characteristics such as gender; Co) college experiences such as year of enrollment; (c) ratings of reasons for initial community college enrollment such as the college is close to home; and (d) current goals such as preparing for career advancement. The rating items had five-point Liken-type scales (5 being most favorable) and two scale anchors (1 = Not at all Important and 5 = Extremely Important).

Pilot Study

A pilot group (N=35) similar to the actual study participants completed the survey instrument twice at a two-week intervals. The criterion for retaining a survey item was a coefficient of stability of r=.60, the minimal reliability coefficient recommended by Nunnally and Bernstein (1994) for use in statistical analysis. The mean item coefficients of stability for the subsections of the survey ranged from .82 to .99. The mean coefficient of stability for all items on the instrument was .89, indicating the instrument had excellent by-item and total test-retest reliability (Borg, Gall, & Gall, 1993; Nunnally & Bernstein, 1994).

Data Analysis

The procedure for examining associations between nominal scaled variables was chi-square analysis. The procedure used to test for significant mean differences between the completer and noncompleter groups was the independent sample t test, which was appropriate because the data met the homogeneity-of-variances assumption for this procedure. Pearson product-moment correlations served to assess associations between interval scaled variables (for example, age and credit hours), and point-biserial correlations served to examine relationships between a naturally dichotomous variable such as reverse transfer status and an interval scaled variable such as age.

Results

A total of 882 individuals responded to the survey questionnaire, yielding a response rate of 63.5%. The survey respondents included 148 completer transfers (16.8%) and 734 noncompleter transfers (83.2%). The results exceeded the criteria established for this study with respect to response rate (60%), number of completer transfers (n = 100), and number of noncompleter transfer ... (n = 733).

Student Characteristics

The first step in the data analysis was to perform chi-square tests to assess the bivariate relationships between the nominal scaled variables noted in Table 1. There was no association between reverse transfer status (completer, noncompleter) and gender (male, female): [chi square] = 2.53, p [is greater than] .05. There was a statistically significant association between reverse transfer status (completer, noncompleter) and race (White, African American, Asian American, Hispanic American, Native American): [chi square] = 12.75, p [is less than] .05. However, the expected frequencies of the cells for African Americans, Asian Americans, and Hispanic Americans failed to meet the minimum (n=5) required for this procedure (Hays, 1988, p. 781). Accordingly, subsequent analysis relative to race involved only the White and African American groups and included the Bonferroni correction (Keppel, 1991, p. 164) to control for family-wise error rate. Results indicated African Americans had a significantly larger proportion of noncompleter transfers (94.8%) than did Whites (82.2%): [chi square] = 8.01, p [is less than] .0125.

Similar analyses detected a:a association between marital status and reverse transfer status. These analyses indicated that the proportion of noncompleters among married respondents (78.4%) was significantly lower than the proportion of noncompleters among single students (87.5%). Further analyses also detected a significant relationship between reverse transfer status and employment status (full-time, part-time, unemployed): [chi square] = 8.06, p [is less than] .0125. The significant difference relative to employment status was within the noncompleter reverse transfer group. The proportion of full-time employed noncompleters (20.6%) was significantly greater than the proportion of unemployed noncompleters (12.6%).

The second step in the data analysis was to conduct independent sample t tests to assess the group (completer, noncompleter) mean differences with respect to the interval scaled data reported in Table 1. The results for the t tests appear in Table 2. When compared to the completer reverse transfers, the noncompleter reverse transfers were significantly younger, had fewer dependent children, enrolled for more credit hours, completed more credit hours, and earned lower grade-point averages.

Table 2 Group Means and Results of Independent Sample t Tests for Student Characteristics (N = 882)

                  Noncompleters   Completers
Variable                M             M

Age                   29.3           37.5       9.4(**)
Dependents             0.5            0.8       2.8(*)
Credit hours           9.0            5.4      -9.5(**)
Hours completed       32.5           23.1      -3.9(*)
GPA                    3.2            3.5       5.1(*)

(*) p < .05

(**) p < .001

The third step in the data analysis was a correlation analysis (see Table 3). The data in Table 3 reflect bivariate relationships only and indicate the degree of strength of the linear relationship between two variables. To assess the degree of strength of the bivariate relationships detected in this analysis, the researchers adopted criteria developed by Cohen (1988) for assessing effect sizes associated with correlation analysis. The effect size criteria established by Cohen (pp. 79-81) are small (r=. 10), medium (r= .30), and large (r= .50).

Table 3 Correlation Matrix for Student Characteristics (N = 882)

Variable                1         2         3

1. Transfer status   1.00
2. Age               -.30(*)   1.00
3. Dependents        -.10(*)    .38(*)   1.00
4. Credit hours       .31(*)   -.36(*)   -.17(*)
5. Hours completed    .14(*)    .04       .02
6. GPA               -.18(*)    .27(*)    .17(*)

Variable                4         5       6

1. Transfer status
2. Age
3. Dependents
4. Credit hours      1.00
5. Hours completed    .20(*)   1.00
6. GPA               -.24(*)   -.08(*)   1.00

Note. Coefficients involving reverse transfer status (completer = 1, noncompleter = 2) are point-biserial correlations. Other coefficients are Pearson product-moment correlations.

(*) p < .01

With respect to participant age, the correlations in Table 3 indicate (a) the relationship between age and number of dependent children (r= .38) was positive and moderate in magnitude; (b) the relationship between age and credit hours enrolled (r=-.36) was negative and moderate in magnitude; and (c) the relationship between age and GPA (r=.27) was positive and small in magnitude. With respect to number of dependent children, these additional bivariate relationships were detected: (a) the association between number of dependent children and credit hours enrolled (r=.17)was negative and small in magnitude; and (b) the association between number of dependent children and GPA (r=.17) was positive and small in magnitude. Also, the relationship between credit hours enrolled and credit hours completed (r=.20) was positive and small in magnitude, and the relationship between credit hours enrolled and GPA (r=-.24) was negative and small in magnitude. Finally, the association between credit hours completed and GPA (r=-.08) was negative and small in magnitude. The correlations between reverse transfer status and the other variables in Table 3, although statistically significant in all cases, did not render information that was new given the t-test results reported about these same variables in Table 2.

College Experiences

The survey items relating to college experiences revealed new information about the degree completion pattern of completer reverse transfers. By definition, 100% of the completer reverse transfers in this study had earned a baccalaureate degree. However, many completer reverse transfers had earned a second college degree, such as the associate's degree (3.4%), the master's degree (23.6%), or a professional degree (2.7%). Almost 30% of the completer participants had earned two degrees prior to enrolling at a community college.

Two survey items queried the participants regarding programs of study. In response to these items, 57.4% of the study participants indicated they were not yet enrolled in a degree program, whereas 42.6% indicated they were enrolled in a specific program. Among students admitted to a program, the highest program frequencies were (a) general studies or academic transfer (28.4%), (b) nursing (18.5%), (c) business technology (11.8%), (d) computer information systems (9.6%), (e) engineering technology (5.9%), (f) medical or dental technology (4.5%), and (g) office administration (3.7%). There was one significant difference between the completer reverse transfers and the noncompleter reverse transfers: Whereas 31.4% of the noncompleters were pursuing a program described as general studies or academic transfer, only 4.8% of the completer reverse transfers were pursuing this program.

One survey item captured the respondents' future degree aspirations. The degree aspirations indicated were (a) complete a baccalaureate degree (58.6%), (b) complete an associate's degree (39.9%), (c) complete a master's degree (32.8%), (d) complete a doctoral degree (6.2%), and (e) complete a professional degree (4.6%), such as the M.D., the D.D.S., or the J.D. Future degree aspirations were highest among the noncompleter reverse transfers, with 88.6% indicating a desire to complete an associate's degree and 94.8% indicating a desire to complete a bachelor's degree. Among the completer reverse transfers, 11.4% indicated a desire to complete an associate's degree and 5.2% indicated a desire to complete a bachelor's degree.

Reasons and Goals for College Attendance

The final sections of the survey contained 23 items with five-point Likert-type scales (5 being most favorable) that measured participant ratings of importance for both reasons for initial community college enrollment (17 items) and for current goals (6 items). Analysis of these items included descriptive statistics and group comparisons (completers, noncompleters). The descriptive analysis for the 23 items revealed 13 items with mean scores greater than or equal to 3.5. The researchers adopted a rating of 3.5 as the criterion for classifying an item as important to the respondents because 3.5 falls above the midpoint on the five-point scales used to capture participant responses.

The items rated as important for initial community college enrollment included (a) college is low cost, (b) courses are offered at convenient locations, (c) college has quality instruction, (d) courses are scheduled at convenient times, (e) college has a good reputation, (f) college is close to home, and (g) improve GPA. The items rated as important current goals were (a) upgrade skills, (b) learn new skills, (c) prepare for career advancement, (d) complete courses for academic transfer, (e) complete courses for personal growth, and (f) complete an associate's degree. The items rated as important appear in rank order by category in Table 4.

Table 4 Rank Order of Reasons for Initial Enrollment and Current Goals Rated as Important (N = 882)

Survey Item                               M

Reasons for Initial Enrollment

Low cost                                 4.12
Convenient course locations              4.01
Quality instruction                      3.95
Convenient course times                  3.93
Good reputation                          3.92
Close to home                            3.80
Improve GPA                              3.59

Current Goals

Upgrade skills                           4.06
Learn new skills                         4.03
Prepare for career advancement           3.99
Complete courses for academic transfer   3.97
Complete courses for personal growth     3.77
Complete associate's degree              3.58

Note. The criterion for including an item as important was a rating of 3.5 on a 5-point scale.

Group comparisons for the 17 items related to reasons for initial college enrollment revealed significant differences between the completer and noncompleter groups. Mean ratings for completer reverse transfers were higher than mean ratings for noncompleter reverse transfers for five items: (a) obtain training for current job (t=2.8, p [is less than] .05), (b) acquire skills for career change (t=4.81, p [is less than] .01), (c) courses are scheduled at convenient times (t-2.89, p [is less than] .05), (d) college is close to home (t-2.4, p [is less than] .05), and (e) college is close to work site (t=5.72, p [is less than] .01).

Mean ratings for noncompleter transfers were higher than mean ratings for completer transfers for eight items: (a) prepare for academic transfer (t=-6.1, p [is less than] .001), (b) increase self confidence (t=-2.7, p [is less than] .05), (c) improve GPA (t=-8.0, p [is less than] .001), (d) improve basic skills (t=-7.0, p [is less than] .001), (e) college has minimal admissions requirements (t=-2.2, p [is less than] .05), (f) college has a good reputation (t=-2.1, p [is less than] .05), (g) complete an associate degree (t=-7.2, p [is less than] .001), and (h) complete courses for academic transfer (t=-6.7, p [is less than] .001),

With respect to the six survey items related to current goals, the two groups of respondents combined rated all items as having a high degree of importance (criterion = 3.5 mean importance rating). Results for the group comparisons for current goals appear in Table 5. Predictably, results indicated noncompleters gave significantly higher ratings than did completers for completing an associate's degree and for completing courses for academic transfer. Surprisingly, even the completer reverse transfers had some interest in earning an associate's degree (M=2.67) and completing courses for academic transfer (M=3.29). These results are consistent with the ambitious degree aspirations reported earlier for noncompleters.

Table 5 Group Means and Results of Independent Sample t Tests for Current Goals (N=882)

                     Noncompleters   Completers
Variable                   M             M            t

Associate's degree       3.75           2.67      -7.22(**)
Academic transfer        4.11           3.29      -6.73(**)
Personal growth          3.76           3.79       0.25
Career advancement       3.98           4.04       0.52
Upgrade skills           4.06           4.04      -0.18
Learn new skills         4.04           4.04        .00

(**) p < .001

Discussion

The survey results obtained in this investigation allowed the researchers to accomplish the research goals, which were to (a) access reliable data about reverse transfers across all institutions within a statewide community college system and (b) develop a profile of reverse transfers to guide implementation of student services, programs, and advocacy. The sections that follow discuss the derived reverse transfer student profile in a student services context, and highlight possible implications of the study's findings for student recruitment and retention, curriculum and instruction, and future reverse transfer research.

Student Profile

The reverse transfers studied in this investigation conform to a nontraditional student profile. Study participants were older (M=30.7 years) than traditional students and had an average of one dependent child. The participants were predominately female (66.4%) and White (88.2%). Also, a large proportion of the participants were married (38.2%) and working (79.2%) while attending college (47.7% full-time, 32.0% part-time). Academically, the study participants were high performers (mean GPA-3.24).

The profile developed suggests reverse transfers need services and programs designed for students who work (evening and weekend classes, flexible class locations) or have families (day care, family counseling). Student orientation and student information programs targeted for reverse transfers should be structured for individuals spending a minimal amount of time on campus. Because they have heavy off-campus responsibilities, reverse transfers appear to require more orientation about available student services than do traditional students. Accordingly, reverse transfers would likely benefit from comprehensive direct mail programs that inform commuter students about academic programs, campus resources, student organizations, and campus facilities such as parking and computer labs. Also, because of their limited on-campus presence, reverse transfers would benefit from systematic efforts to create more personal interaction with faculty and staff.

The data about participant reasons for initial enrollment at a community college (see Table 4) appear to be an encouraging result for the faculty and administrators of community colleges in Kentucky. The ratings in Table 4 indicate the participants enrolled at a community college for reasons such as low cost, convenient location, and convenient class scheduling. These services are associated with the traditional community college mission to serve place-bound students with limited financial resources (Cohen & Brawer, 1996). Also, two quality indicators received high ratings. The participants indicated that both the quality of instruction and the good reputation of the college were important reasons for initial community college enrollment. All of the above items suggest the Kentucky community colleges are serving students well in important service areas. Ratings were also high for "improve GPA," which suggests the need for research about the past academic performance of reverse transfers. Did the reverse transfers in this study perform poorly at another institution, or did other factors influence the participant emphasis on GPA improvement? The current goals items displayed in Table 4 may offer a partial answer to these questions.

Two of the six current goals in Table 4 (complete courses for academic transfer, complete associate's degree) relate to academic performance. Regardless of their past performance, the study participants were focused on doing well academically. The participants were also focused on enhancing their careers as evidenced by the ratings given for upgrading skills, learning new skills, and preparing for career advancement. Finally, the participants appeared to have an interest in general self-improvement as evidenced by the high ratings for completing courses for personal growth.

Student Recruitment and Retention

Reverse transfers constituted 11.1% of the Kentucky community college student population, a figure consistent with the average (12%) used in this study. In fact, only two colleges in Kentucky had enrollments larger than the overall reverse transfer population, which is a further indication of the importance reverse transfers can have relative to the total enrollment of a statewide system.

Reverse transfers may present an opportunity in terms of student recruitment and retention. The study data suggest reverse transfers are a desirable addition to the student body in terms of academic performance. The mean GPA of the study participants placed reverse transfers well within the top half of students in the general population (University of Kentucky Community College System, 1996). Noncompleters earned a mean GPA of 3.2, and completers earned a mean GPA of 3.5.

The study participants reported exceptionally high future educational attainment aspirations. Likewise, study data indicate reverse transfers have both the desire and the ability to earn the degrees and academic credentials to which they aspire. These findings suggest that targeting reverse transfers for recruitment may have a positive impact on both student retention rates and on program completion rates. Given the high academic performance detected among reverse transfers, it appears, also, that recruiting more reverse transfers would put minimal stress on existing academic advising resources. That is, targeting reverse transfers for recruitment may offer community colleges the opportunity to achieve higher student enrollments and higher student retention rates with a minimal increase in academic advisement costs. However, some policy makers could argue that targeting reverse transfers might achieve the above outcomes while failing to consider the opportunity costs associated with college access for other student groups. It would appear that these conflicting perspectives can be reconciled only by conducting a cost-benefit analysis in particular instances when college officials are considering changes in existing student recruitment policies.

Curriculum and Instruction

With respect to students' reasons for initial enrollment and current goals, study findings appear to have positive implications for curriculum and instruction. Many of the reasons given by reverse transfers for attending a community college (for example, low cost, location close to home, convenient course scheduling) constitute known competitive advantages for community colleges in comparison with four-year colleges (Cohen & Brawer, 1996). Regardless of whether they are completers or noncompleters, reverse transfers are highly focused on obtaining and upgrading skills that are directly related to current and future job requirements. The completer and noncompleter transfers in this study rated the following survey items most favorably with respect to current goals: (a) career advancement, (b) upgrade skills, and (c) learn new skills. This focus on job-related needs may be regarded as another opportunity for community colleges in that they historically have been highly proficient at delivering education with immediate career advancement or job enrichment impact.

Finally, in comparison with four-year colleges, community colleges have various competitive advantages relative to designing and delivering education with near-term job application (Cohen & Brawer, 1996) including (a) flexible scheduling, (b) flexible course locations (including the work site), (c) experience in delivering custom-designed education for organizations of all types, and (d) governance structures that permit short turnaround times for curricular changes to meet student needs. These advantages can be leveraged to attract reverse transfers in situations where community colleges can identify pools of individuals within occupations who need college-level instruction for career advancement or to avoid skill obsolescence.

Conclusion

Within the context of the SPAR Model (Jacoby & Girrell, 1981) that framed this investigation, the study results yielded new knowledge about a nontraditional group of community college students (reverse transfers). Although much remains to be discovered about the motivations of reverse transfer students, it is hoped that the results of this study will stimulate additional inquiry about this intriguing student group. Future research should focus on two issues: (a) generating information about reverse transfers within other statewide community college systems and (b) exploring additional factors that contribute to the reverse transfer phenomenon.

References

American Association of Community Colleges. (1997). AACC 1996-97 annual. Washington, DC: Author.

American Council on Education. (19!)8). Fact book on higher education: 1997 edition. Phoenix, AZ: The Oryx Press.

Babbie, E. (1992). The practice of social research (6th ed.). Belmont, CA: Wadsworth.

Barr, M. L., & Associates (Eds.). (1993). The handbook of student affairs administration. San Francisco: Jossey-Bass.

Borg, W. R., Gall, J. P., & Gall, M. D. (1993). Applying educational research: A practical guide (3rd ed.). New York: Longman.

Brimm, J., & Achilles, C. M. (1976) The reverse transfer student: A growing factor in higher education. Research in Higher Education, 4, 355-360.

Clark, B. R. (1960). The open-door college: A case study. New York: McGrawHill.

Clark, D. O. (1982). Reverse Transfers, laterals, double transfers, other regular irregulars. College and University, 57, 166-168.

Cohen, A. M., & Brawer, F. B. (1996). The American community college (3rd ed.). San Francisco: Jossey-Bass.

Cohen, A. M., Palmer, J. C., & Zwemer, K. D.. (1986). Key resources on community colleges. San Francisco: Jossey- Bass.

Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). New Jersey: Lawrence Erlbaum Associates.

Dillman, D. A. (1978). Mail and telephone surveys: The total design method. New York: John Wiley and Sons.

Fowler, F. J. (1988). Survey research methods (2nd ed.). Newbury Park, CA: Sage Publications.

Gall, M. D., Borg, W. R., & Gall, J. P. (1986). Educational research: An introduction (6th ed.). New York: Longman.

Hays, W. L. (1988). Statistics (4th ed.). Chicago: Holt, Rinehart and Winston.

Hogan, R. R. (1986). An update on reverse transfers to two-year colleges. Community/Junior College Quarterly, 10, 295-306.

Jacoby, B., & Girrell, K. W. (1981). A model for improving services and programs for commuter students. NASPA Journal, 18, 38-41.

Kajstura, A., & Keim, M. C. (1992). Reverse transfer students in Illinois community colleges. Community College Review, 20(2), 39-44.

Keppel, G.. (1991). Design and analysis: A researchers' handbook (3rd ed.). Englewood Cliffs, NJ: Prentice-Hall.

Klepper, D. F. (1990). A descriptive analysis of completer reverse transfer students at a Virginia community College. Dissertations Abstracts International, 51(11), 3971A. (University Microfilms No. AAC 9407654).

Kuznik, A. E., Macey, E. J., & Anderson, D. D. (1974). Reverse transfers: A survey. Community and Junior College Journal, 44(4), 25-27.

Lee, G.. (1975). Reverse transfer: The "retread function" of community colleges. Paper presented at the annual meeting of the American Educational Research Association, Washington, DC. (ERIC Document Reproduction Service No. ED 104 471)

Mitchell, G. N., & Grafton, C. L. (1985). Comparative study of reverse transfer, lateral transfer, and first-time community college students. Community/Junior College Quarterly, 9, 273-280.

Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory (3rd ed.). New York: McGraw-Hill.

Renkiewicz, N. K., Hirsch, P. M., Drummond, M. E., & Mitchell, G. E. (1982). The reverse transfer student: An emerging population. Sacramento, CA: Northern California Community College Research Group.

Ross, R. (1982). The reverse transfer phenomenon at Piedmont Virginia Community College (Research Report No. 3-83). (ERIC Document Reproduction Service No. ED 216 758).

Swedler, J. A. (1983). Reverse transfers who return to a university: An analysis of their academic progress. Community/Junior College Research Quarterly, 7, 131-137.

University of Kentucky Community College System. (1996). UK community college system fact book. Frankfort, KY: Author.

Paul A. Winter is a professor in the Department of Administration and Higher Administration, School of Education, at the University of Louisville in Louisville, Kentucky. ([e-mail])

Michael R. Harris is director of Recruiting, Publications, and Financial Aid for the Speed Scientific School at the University of Louisville in Louisville, Kentucky. ([e-mail])