Colleges and universities are facing more intense pressures to demonstrate their accountability, effectiveness, and efficiency (Wolkwein, 2003). Under this pressure, higher education institutions are expected to perform, to document their performance, and to be accountable for producing return on taxpayer and student investment (McCleeney, 1998, cited in Alfred, Ewell, Hudgins, & McClenney, 1999). One of the results of these developments is to broaden the scope of assessment of institutional effectiveness. Student retention has been suggested as one of the major performance indicators for community colleges nation wide (Alfred, Ewell, Hudgins, & McClenney, 1999). Although student retention has been investigated extensively in the past decades, the focus was primarily on four-year institutions. Research on the most diverse student populations in two-year colleges is relatively sparse, where attrition is the highest (Grimes & Antworth, 1996). The profile of the typical community college student, a person who has a full- or part-time job, lives off campus, and is taking classes on a part-time basis, makes retention a particular challenge for community colleges (Bonham & Luckie, 1993). Unlike residential university students, community college students are constantly balancing the social and academic demands of the college campus with the responsibilities of family and work. Therefore, retention issues in community colleges are different from those in 4-year universities.
To better understand unique community college student retention, this study is designed to identify and describe characteristics of community college student withdrawals and reasons for students’ decision to leave. Student demographic information was analyzed and compared. Additionally, a surveys was also sent out to those who left to determine the reasons for leaving. It is hoped that this research will yield some important information that can serve as the foundation for any effort to improve student retention and success at community colleges.
Review of Literature
The results of many studies suggested that retention is a complex issue that involves different factors (Umoh, Eddy, & Spaulding, 1994). Recognized variables include various student characteristics and student-institutional interaction, academic aptitude and performance, level of aspiration and motivation, institutional type, student services offered, using of student counseling, and student involvement-plus the development of a sense of belonging or degree of fit that results from student and institution interactions (Beal & Noel, 1980; Lenning, Beal, & Sauer, 1980; Turner & Berry, 2000). Furthermore, retention, student satisfaction, and student success appeared to improve when retention efforts were geared toward integrating the students’ total educational experience (Umoh et al., 1994).
Researchers also believed that variables related to prior academic performance, socioeconomic background, and educational goals may influence student retention. For example, Kinnick and Kempner (1988) found a significant association between student retention and the following variables: high school GPA, socioeconomic background, parental income, type of college first attended (two- or four-year), educational aspiration, and high school preparatory programs. Those more likely to complete a bachelor’s degree were those with a higher high school GPA, higher socioeconomic status, initially attended a four-year institution, higher degree aspirations, and completed a college preparatory program in high school (Kinnick & Kempner, 1988).
The campus environment can also affect students’ decisions to stay in or drop out of school, particularly developmental education students. Studies indicated that developmental education students who withdrew from college often rate the presence of a hostile racial climate as an important reason for their withdrawal (Pascarella & Terenzini, 1991). Observers reported that faculty, especially in senior colleges and research institutions, often had negative attitudes toward remedial students and were poorly prepared to teach them (Gross, 1981). Increasing the presence of representative group faculty members on campus has been recommended as a fundamental step toward improving campus environments. As early as the 1970s, the shortage of African-American and Hispanic instructors at two-year colleges was noted as a particular problem (Olivas, 1979). In a recent study, Braxton, Milem, and Sullivan (2000) found that faculty classroom behaviors played a role in the student departure process. The authors indicated that faculty classroom behaviors in general and active learning in particular may constitute an empirically reliable source of influence on social integration, subsequent institutional commitment, and departure decisions (Braxton et al., 2000).
Many researchers have examined the relationship between retention and student demographic characteristics. Glass and Garrett (1995) found that retention and college GPA were not related to age, gender, race, employment status, college major, or college attended. However, Tinto (1993) pointed out that older students were more likely to drop out because typical adult students were more likely to be married, to have children at home, to live off campus, and/or to be employed while attending college. Therefore, they were more likely to encounter greater problems in finding adequate on-campus time to study in order to meet the minimum academic standards of the institution (Tinto, 1993). Furthermore, many of these students either take one or two courses at a time or leave for a semester or two and then return. Because the average community college student is generally older than students attending four-year universities, they are more likely to fit the profile of the adult student described by Tinto.
Several researchers have examined retention issues specific to community colleges. Voorhees’s 1986 study employed a log-linear modeling approach to explore the conceptual relationships between community college student persistence and several variables, including student demographics, purpose for enrolling, intentions to return, frequency of informal interaction with faculty, and satisfaction with the institution in general. Results of this study revealed that full-time female students had greater persistence rates than their male counterparts, while GPA, number of hours spent studying each week, and frequency of interaction with faculty, had independent effects on student persistence. Daniels (1990) also found students’ academic goals and intentions tended to significantly influence retention. In other words, the higher the level of one’s educational goals, the greater the likelihood of college completion. This was because the goal of occupational attainment became the motivating force for undertaking and completing a particular academic degree program (Tinto, 1993).
Research Methods
The population of this study was defined as those who dropped out of San Diego Community College District during fall semester of 2000 and those who did not return for the following spring semester of 2001. Three types of student withdrawals were identified based on withdrawing time frames: 1) students who left before the semester started: those who filed an application for fall 2000 semester, but did not enroll in any classes (n = 5,459), 2) students who left during the semester: those who withdrew from all classes during fall 2000 semester (n = 7,481), and 3) non-persisters: those who completed fall semester of 2000 but did not return in spring semester of 2001 (n = 10,968). Survey questionnaires were sent out to random samples of these three types of student withdrawals to assess their reasons for leaving. Surveys were administered in fall 2000 and spring 2001 with an overall response rate of 10.7% (see Table 1). Due to the limited funding, no follow-up efforts were made to non-respondents.
Table 1
Survey Response Rates
|
|
Surveys mailed
|
Surveys returned
|
Response rate
|
|
Left before semester started
|
2,004
|
193
|
9.6%
|
|
Left during semester
|
3,501
|
413
|
11.8%
|
|
Non-persister
|
5,000
|
515
|
10.3%
|
|
Total
|
10,505
|
1,121
|
10.7%
|
The following two research questions were developed to guide this study:
1. What are the demographic differences among students who withdrew at three different time frames and to what extent do they differ from the general student population?
2. What factors influenced students’ decision of withdrawing and do these factors differ among students who withdrew at different time frames?
Results
Descriptive statistics were used to analyze the demographic information and survey results. Results are presented below:
Research Question 1. What are the demographic differences among students who withdrew at three different time frames and to what extent do they differ from the general student population?
Table 2 presents the gender, ethnicity, and age distributions of three types of student withdrawals. Characteristics about the student population (N=45,498) are also included as a basis for comparison. As can be seen in Table 2, a slightly higher percentage of male students (50.7%) left before the semester started when compared to the general student population (48.1%). For students who left during the semester, the proportion of females was slightly higher (54.1%) than the general population (51.7%). Gender distribution of non-persisters was similar to general student population.
Table 2
Student Demographics by Type of Withdrawals
|
Demographics
|
General population
|
Left before semester started
|
Left during semester
|
Non-persister
|
|
|
#
|
%
|
#
|
%
|
#
|
%
|
#
|
%
|
|
Gender
|
|
|
|
|
|
|
|
|
|
Female
|
23,502
|
51.7%
|
2,702
|
49.2%
|
4,049
|
54.1%
|
5,731
|
52.3%
|
|
Male
|
21,896
|
48.1%
|
2,786
|
50.7%
|
3,420
|
45.7%
|
5,224
|
47.6%
|
|
Not reported
|
100
|
0.2%
|
7
|
0.1%
|
12
|
0.2%
|
13
|
0.1%
|
|
|
|
|
|
|
|
|
|
|
|
Ethnicity
|
|
|
|
|
|
|
|
|
|
American Indian
|
524
|
1.2%
|
86
|
1.6%
|
92
|
1.2%
|
121
|
1.1%
|
|
Asian
|
4,945
|
10.9%
|
419
|
7.6%
|
742
|
9.9%
|
1,111
|
10.1%
|
|
African American
|
4,470
|
9.8%
|
821
|
14.9%
|
846
|
11.3%
|
1,012
|
9.2%
|
|
White
|
2,0012
|
44.0%
|
2,194
|
39.9%
|
3,340
|
44.6%
|
4,963
|
45.2%
|
|
Latino
|
7,814
|
17.2%
|
1,084
|
19.7%
|
1,248
|
16.7%
|
1,832
|
16.7%
|
|
Filipino
|
2,830
|
6.2%
|
251
|
4.6%
|
443
|
5.9%
|
683
|
6.2%
|
|
Pac Islander
|
431
|
0.9%
|
60
|
1.1%
|
73
|
1.0%
|
94
|
0.9%
|
|
Other
|
1,976
|
4.3%
|
358
|
6.5%
|
313
|
4.2%
|
483
|
4.4%
|
|
Not reported
|
2,496
|
5.5%
|
222
|
4.0%
|
384
|
5.1%
|
669
|
6.1%
|
|
|
|
|
|
|
|
|
|
|
|
Age
|
|
|
|
|
|
|
|
|
|
17 or under
|
2,234
|
4.9%
|
290
|
5.3%
|
96
|
1.3%
|
634
|
5.8%
|
|
18 24
|
21,030
|
46.6%
|
2,838
|
51.6%
|
3,040
|
40.7%
|
4,961
|
45.3%
|
|
25 34
|
12,120
|
26.6%
|
1,324
|
24.0%
|
2,490
|
33.3%
|
2,954
|
27.0%
|
|
35 or over
|
10,114
|
22.2%
|
1,043
|
19.0%
|
1,855
|
24.8%
|
2,419
|
22.1%
|
|
|
|
|
|
|
|
|
|
|
|
Total
|
45,498
|
100%
|
5,495
|
100%
|
7,481
|
100%
|
10,968
|
100%
|
With regard to ethnicity, it should be noted that of the students who left before the semester started, there was a higher proportion of African Americans (14.9%) and Latinos (19.7%) as compared to the proportions in total student population. The ethnic distribution of the students who left during the semester and non-persisters was similar to the general population.
Finally, when compared to the general population, students who left before the semester started had a higher proportion of students between the ages of 18 and 24 (51.6% vs. 46.6% in the population). Thus, students who left before school started tended to be younger. However, for students who withdrew during the semester, they tended to be older since there was a higher proportion of them between the ages of 25-34 (33.3% vs. 26.6% in the population). No notable difference was found between non-persisters and student population.
Table 3
High School GPA by Type of Withdrawals
|
High School GAP
|
General population
|
Left before semester started
|
Left during semester
|
Non-persister
|
|
|
#
|
%
|
#
|
%
|
#
|
%
|
#
|
%
|
|
< 2.0
|
402
|
0.9%
|
70
|
1.3%
|
80
|
1.0%
|
90
|
0.8%
|
|
2.0-2.5
|
5,079
|
11.2%
|
744
|
13.5%
|
935
|
12.5%
|
1,053
|
9.6%
|
|
2.6-2.9
|
10,297
|
22.6%
|
1,280
|
23.3%
|
1,689
|
22.6%
|
2,242
|
20.4%
|
|
3.0-3.5
|
10,219
|
22.5%
|
1,069
|
19.5%
|
1,539
|
20.6%
|
2,657
|
24.2%
|
|
> 3.5
|
6,670
|
14.7%
|
539
|
9.8%
|
842
|
11.3%
|
2,004
|
18.3%
|
|
Not reported
|
12,831
|
28.2%
|
1,793
|
32.6%
|
2,396
|
32.0%
|
2,922
|
26.6%
|
|
|
|
|
|
|
|
|
|
|
|
Total
|
45,498
|
100%
|
5,495
|
100%
|
7,481
|
100%
|
10,968
|
100%
|
Self-reported high school GPA information is summarized in Table 3.Data show that students who left before the semester started and those who withdrew during the semester had a smaller percent of them with a high school GPA over 3.5, as compared to the population.Non-persisters had a higher proportion with high school GPAs over 3.5 than the general population (18.3% vs. 14.7% in the population).
Table 4
Educational Objective by Type of Withdrawals
|
Educational Objective
|
General population
|
Left before semester started
|
Left during semester
|
Non-persister
|
|
|
#
|
%
|
#
|
%
|
#
|
%
|
#
|
%
|
|
Transfer
|
20,858
|
45.9%
|
2,181
|
39.7%
|
3,300
|
44.1%
|
4,672
|
42.6%
|
|
Obtain AA/AS/Certificate No Transfer
|
3,015
|
6.7%
|
474
|
8.6%
|
492
|
6.6%
|
656
|
6.0%
|
|
Career Related
|
8,141
|
17.9%
|
1,074
|
19.5%
|
1,422
|
19.0%
|
2,222
|
20.2%
|
|
Other
|
3,188
|
7.0%
|
548
|
10.0%
|
641
|
8.6%
|
977
|
8.9%
|
|
Undecided
|
9,142
|
20.1%
|
1,116
|
20.3%
|
1,428
|
19.1%
|
2,098
|
19.1%
|
|
Not reported
|
1,154
|
2.5%
|
102
|
1.9%
|
198
|
2.6%
|
343
|
3.1%
|
|
|
|
|
|
|
|
|
|
|
|
Total
|
45,498
|
100.0%
|
5,495
|
100.0%
|
7,481
|
100.0%
|
10,968
|
100.0%
|
Educational objective was identified to be related to student retention in previous research; therefore, this variable was also examined in this study (See Table 4). Results show that the educational objective for 45.9% of the general population was to transfer.This rate is similar for students who withdrew during the semester (44.1%), but slightly lower for non-persisters (42.6%) and even lower for those who left before school started (39.7%). It is also noted that non-persisters had a slightly higher proportion of students who reported their educational objective was career related.
Table 5
Annual Family Income by Type of Withdrawals
|
Annual Income
|
General population
|
Left before semester started
|
Left during semester
|
Non-persister
|
|
|
#
|
%
|
#
|
%
|
#
|
%
|
#
|
%
|
|
$0-2,999
|
1,137
|
2.5%
|
203
|
3.7%
|
215
|
2.9%
|
240
|
2.2%
|
|
$3,000-5,999
|
683
|
1.5%
|
82
|
1.5%
|
145
|
1.9%
|
100
|
0.9%
|
|
$6,000-9,999
|
2,000
|
4.4%
|
237
|
4.3%
|
390
|
5.2%
|
364
|
3.3%
|
|
$10,000-14,999
|
4,076
|
9.0%
|
594
|
10.8%
|
738
|
9.9%
|
823
|
7.5%
|
|
$15,000-20,999
|
4,370
|
9.6%
|
606
|
11.0%
|
830
|
11.1%
|
1,047
|
9.5%
|
|
$21,000-26,999
|
2,821
|
6.2%
|
355
|
6.5%
|
506
|
6.8%
|
666
|
6.1%
|
|
$27,000-32,999
|
2,691
|
5.9%
|
343
|
6.2%
|
492
|
6.6%
|
712
|
6.5%
|
|
$33,000 or over
|
10,417
|
22.9%
|
1,035
|
18.8%
|
1,548
|
20.7%
|
2,633
|
24.0%
|
|
Not reported
|
17,303
|
38.0%
|
2,040
|
37.1%
|
2,617
|
35.0%
|
4,383
|
40.0%
|
|
|
|
|
|
|
|
|
|
|
|
Total
|
45,498
|
100.0%
|
5,495
|
100.0%
|
7,481
|
100.0%
|
10,968
|
100.0%
|
Table 5 reports family income by withdrawal types.Results reveal that there was a slightly smaller percent of students who left before the semester started had a family income of $33,000 and above (18.8%), when compared to the general population (22.9%). The other two groups were very similar to the student population.
|
Working Hours/Week
|
General population
|
Left before semester started
|
Left during semester
|
Non-persister
|
|
|
#
|
%
|
#
|
%
|
#
|
%
|
#
|
%
|
|
Not working
|
10,654
|
23.4%
|
1,283
|
23.3%
|
1,529
|
20.4%
|
2,620
|
23.9%
|
|
Part-Time (1 39 hours/week)
|
20,456
|
45.0%
|
2,501
|
45.5%
|
3,287
|
44.0%
|
4,526
|
41.2%
|
|
Full-Time (40 hours or more/week)
|
13,819
|
30.4%
|
1,636
|
29.8%
|
2,551
|
34.1%
|
3,721
|
33.9%
|
|
Not reported
|
569
|
1.3%
|
75
|
1.4%
|
114
|
1.5%
|
101
|
0.9%
|
|
|
|
|
|
|
|
|
|
|
|
Total
|
45,498
|
100.0%
|
5,495
|
100.0%
|
7,481
|
100.0%
|
10,968
|
100.0%
|
One of the major characteristics of community college students is that most work part-time or full-time.Student employment information is summarized in Table 6.As expected, a majority of the general student population is working either part-time (45.0%) or full-time (30.4%).In comparison with the student population, students who withdrew during the semester and non-persisters had slightly higher proportions of students working full-time (34.1% and 33.9% respectively).However, employment hours for students who left before semester started were very similar to the population.
Table 7
Enrollment Status by Type of Withdrawals
|
Enrollment Status
|
General population
|
Left before semester started
|
Left during semester
|
Non-persister
|
|
|
#
|
%
|
#
|
%
|
#
|
%
|
#
|
%
|
|
Current High School Student
|
1,465
|
3.2%
|
172
|
3.1%
|
57
|
0.8%
|
379
|
3.5%
|
|
First-Time Student
|
4,569
|
10.0%
|
2,296
|
41.8%
|
723
|
9.7%
|
1,049
|
9.6%
|
|
First-Time Transfer Student
|
6,718
|
14.8%
|
1,850
|
33.7%
|
955
|
12.8%
|
2,847
|
26.0%
|
|
Returning Transfer Student
|
1,933
|
4.2%
|
380
|
6.9%
|
397
|
5.3%
|
671
|
6.1%
|
|
Returning Student
|
3,211
|
7.1%
|
797
|
14.5%
|
737
|
9.9%
|
884
|
8.1%
|
|
Continuing Student
|
27,349
|
60.1%
|
|
|
4,556
|
60.9%
|
5,052
|
46.1%
|
|
Not Reported
|
253
|
0.6%
|
|
|
56
|
0.7%
|
86
|
0.8%
|
|
|
|
|
|
|
|
|
|
|
|
Total
|
45,498
|
100.0%
|
5,495
|
100.0%
|
7,481
|
100.0%
|
10,968
|
100.0%
|
Table 7 shows enrollment status across the different types of withdrawals.Data suggest that first-time freshmen and first-time transfers constituted a higher percentage (75.5%) of the students who left before the semester started (24.8% in the general population).However, of the students who withdrew during the semester, only 22.5% were first-time freshmen or first-time transfers, as compared to 24.8% in the general population.There was a larger percent of first-time transfers among non-persisters (26.0% vs. 14.8% in the population).
|
|
General population
|
Left before semester started
|
Left during semester
|
Non-persister
|
|
Cumulative GPA
|
|
|
|
|
|
|
|
|
|
Mean
|
2.46
|
|
N/A
|
|
1.81
|
|
2.27
|
|
|
Std Deviation
|
1.22
|
|
N/A
|
|
1.48
|
|
1.89
|
|
|
Minimum
|
0
|
|
N/A
|
|
0
|
|
0
|
|
|
Maximum
|
4
|
|
N/A
|
|
4
|
|
4
|
|
|
Cumulative Units
|
|
|
|
|
|
|
|
|
|
Mean
|
26.4
|
|
N/A
|
|
21.5
|
|
21.3
|
|
|
Std Deviation
|
29.3
|
|
N/A
|
|
30.3
|
|
29.3
|
|
|
Minimum
|
0
|
|
N/A
|
|
0
|
|
0
|
|
|
Maximum
|
388
|
|
N/A
|
|
288
|
|
302
|
|
Table 8 shows students’ cumulative college GPA and units earned. Results suggest that students who withdrew during the semester tended to have the lowest GPA (mean =1.81) when compared with the general student population and other groups. Non-persisters also had a slightly lower GPA than that of the general student population. Student who left during the semester and non-persisters earned fewer units than the population.
Research Question 2: What factors influenced students’ decision of withdrawing and do these factors differ among students who withdrew at different time frames?
A survey designed to assess students’ withdrawal reasons was sent to a random sample of students from each of the three types of withdrawal groups described above. In the survey, a list of possible reasons was provided and students were asked to indicate their primary reasons for leaving the college. Survey results are summarized in the following Table 9.
|
|
Left before semester started (n =193)
|
Left during semester
(n=413)
|
Non-persister
(n=515)
|
|
|
|
|
|
|