COLLEGE STUDENTS RETENTION
Introduction
The following research analyzes what the predictors of first-year students retention level at community colleges are. Findings indicate that learning institutions incur more to recruit students than retaining them (Tinto, 1993). The outcome of such measures had been that learning institutions allocate a considerable amount of money for recruiting and other associated expenses such as travel and material used for the recruiting purpose. What had been highlighted as a fallout in the part of these same learning institutions is that what they do to retain the students they recruit after they are enrolled falls short of ameliorating the dropout problem. The suggestion is that with an average attrition rate of 41 from first year to second year and a 34 persistence-to-degree rate, learning institutions should change their focus and examine closely what the effect of the predictors of student retention is.
Hossler (2005) had indicated that most learning establishments whether they are community colleges or universities fail in their effort to study what the root cause of dropping out is, as well as there had been lack of intervention programs and methods that would deal with the unique needs of each institution. It had been proven that it is possible and effective to use data to predict what each institutions ability to retain students is, which also could be used to introduce intervention methods tailor-made to the needs of various students who will find their way to a given learning institution, with their particular characteristics and aspirations. Based on that, this research will focus on an empirical research conducted at a given community college in West Texas.
Literature Review
There had been a sizeable research conducted in the area by various sources. Astin (1993), Pascarella Terenzini (1991) and Tinto (1993) had shown that most institutions spend money on recruitment programs and neglect the student retention aspect, to the point where all of them had suffered the negative effect of such neglect. Hossler (2005) also had shown that there is not enough effort exerted in the part of learning institutions to find out what kind of contribution would retention intervention programs will bring to the fore. Tinto (1993) had the credit for his contribution of the student integration model that suggests that, students progress through stages to become mature students. What contributes to the stages are the kind of academic and social integration that are prevalent and making such integration prevalent is the responsibility of learning establishments, because they highly benefit the students and would boost the retention level of learning establishments.
Bean (1990) had approached the retention problem with psychological model that he called student attrition model that postulates the students background has an influential effect on how students interact with the learning establishments they are joining. According to him, the prevalent environmental variables and the goals or intentions of the students are also influencing and deciding factors. Astin (1991) also had a model called input-environment-outcome model that stipulates the outputs such as the kind of degrees earned, the number of students graduated should be seen in parallel to what he called inputs, such as the ability of students, their gender, age, and what they are majoring in. However, he states that it is not possible to take input and output alone without paying attention to the environment, which is made up of the courses, programs, the available facilities, faculty, and peer groups. There are many more researchers and authors who have studied the student retention predictors, some of whom will be discussed in the main body of the essay.
Methodology
The quantitative study employed had made its focus first time in college (FTIC) students in a community college in Texas, where the student population was around 10,000. The years used for data evaluation were 2001, 2002, 2003, and 2004. There were two variables used for the analysis and the first one was the fall semester of first- year to the same year of spring semester retention level. The second one was first-year fall semester to second-year fall semester. There were many other factors that were taken into consideration through the process and some of them were gender, age, ethnic background, whether students had participated in developmental math, reading, and writing courses or not, if they have taken advantage of the student support service or not in particular the one that cater to special-needs students. The source of financing for the education, participation in Internet courses, the hours involved in the semesters enrolled, number of hours dropped from the first semester, and what kind of educational level the parents have are also among the factors. Other statistical methods such as means, standard deviation, and percentage, including chi-square were part of the assessment process. Other variables employed in the calculation included bivariate correlation predictor, point-beserail correlation coefficients, phi correlation coefficients, in addition to multivariate logistic regression modes were included so that it will be possible to put a handle on any cofounders such as demography etc. It is possible to infer the rest from the tables shown in the appendix.
Findings
The retention level inferred from the participants showed that in the spring the lowest rate was 65.7 while the highest was 70.7 showing that at least one-third of FTIC students enrolled in the fall semester were not attending the same institution in the spring. When looking at the fall-to-fall semester retention level there were not a significant difference although the rate was lower at 45.4 and 49.4 showing that more than half of those enrolled in the fall did not come back in the subsequent semesters. When looking at the correlation of predictors, completing a developmental reading course had finished at the top. Completing the developmental math course, getting financial aid, participating on Internet courses, and using the student support service also have positive correlation to the retention rate. If there had been any negative correlation noticed it was students age and semester hours skipped in the first semester. Other predictors such as the parents education level were not strong indicators of retention.
More findings to highlight might be the multivariate model predicting what takes place between the first-fall and first-spring semesters. None of the variables mentioned as predictors had significance statistically. It is possible to view those that have some effects in Table 4 that shows how the various predictors are affecting the retention rate. Important findings worth mentioning are age in the multivariate model had insignificant magnitude, while mothers education level also did not have a significant influence on the retention rate. However, attending developmental education courses affects the retention level positively. The same applies to those who completed the developmental math course since they had shown a high retention level. Since the same applies to those who attended the developmental reading course, it is possible to conclude that those who attended and completed the various available courses have a better odd in the retention rate.
Discussion
When looking at the retention models used there main focus is highlighting how the students and the learning institution are interacting, while at the same time the models convey the message that students attributes that can influence retention are indispensable. Another worthy factor to take note of is that the research conducted is conclusive of all postsecondary students with the assumption that the age factor will reflect what is acceptable traditionally for college students. However, when looking at community colleges the age mix is different and there are students of older age than the normal average where according to Aslanian (2001) about 60 of the student population could be 25 or older, especially in the two-year community colleges. The racial makeup is also different since there are higher percentage of minority students enrolled in such colleges who are attracted according to Cohen and Brawer (1996) by the ease of access, since there are many programs geared toward minority students, the tuition is kept low, and there is also an open-door policy that makes it easy for minority students to join community colleges. Other endemic characteristics of minority students are enrolling for part time schooling and being from low-income families. Furthermore, community colleges enroll underprepared students when compared to universities that require certain grade attainment, because they do not offer neither developmental education nor remedial courses that will enable the students to achieve the desired level. The figures according to McCabe (2000) are 95 of community colleges offer remedial courses and 41 students entering community colleges and 29 of the overall student population entering higher learning institutions for the first time are undeprepared. In addition Thayer (2000) had shown that the parents of students entering community colleges are different than students entering universities such that what he called first-generation students tend to join community colleges and their retention level is also the lowest. Because of that the method employed had emulated to distinguish university students and community college students by using different variables such as age, because the students at community colleges are older, ethnicity since community colleges are the first entry point into higher education for minorities, the kind of developmental courses that are prevalent since those entering community colleges are for the most part underprepared, and numbers of hours in each semester since as much as two-thirds of the students entering community colleges are studying on a part time basis. The fact that the main drive of community colleges is to avail a 2-year transferable degree, as well as terminal certificate to enhance job skill had also been under consideration. Other factors such as parental education, the financial aid package that is available for the two establishments, what kind of online courses both are availing were part of the consideration.
Based on all these factors, the retention level was highly dependent on the available developmental education that is the major retention contributor. What this meant is students who were underprepared, took and completed the developmental courses that include math, writing, and reading did not only have a higher retention level, but their chance of completing their education was higher. Accordingly, Higbee, Arendale, and Lundell (2005) had estimated that if it were not for the remedial courses at least two million students would have been dropouts in stead of one-third of them attaining an associate degree. Among the three subjects reading was the most important one if the students did not have it right when they join the colleges (Dixon, 1993). When it comes to math, passing the developmental course was a good indicator of retention for those who join fall-to-spring and fall-to-fall semesters, where even taking a partial course without completing it and passing the tests would mean a higher indicator of retention. Hall and Ponton (2005) had argued that the reason why not taking the developmental math without completing it affect the retention level is not clearly known yet. Waycaster (2001) on the other hand had shown that other variables such as advising, counseling, monitoring the effect of the teaching process and the progress attained, if possible smaller classes for paying better attention to the need of each student could also contribute to the retention level of students who have participated in developmental math.
Developmental writing is also a predictor of fall-to-fall semester retention and is not a strong retention indicator to fall-to-spring semester retention where the reason for that according to Crews and Aragon (2007) and a few others is unknown.
Furthermore, Allen and Seaman (2007) had noted that Internet courses are strong predictor of retention especially for nontraditional students at community colleges that has resulted with a need to come up with more effective online courses. The reason why distant learning sources such as Internet courses are becoming popular is students are not finding what they need at convenient times and locations. Although the reason behind why Internet courses contribute to higher student retention is unclear, the study had found out they are beneficial for retention. Participating in the government funded TRIO program with its small number of participants had shown a higher level of retention level when seen from the 397 students who participated that make 3.9 of the overall assessed student population.
Getting financial aid also raises the students retention level since many community college students cite financial problems as their main reason for quitting their education (Zhai and Monzon (2001). Although there are not models developed to show why financial aids add to the student retention level, Lichtenstein (2002) had also shown there is persistence associated with financial aid. As far as first generation students are concerned, the findings had indicated that their retention level is low. However, this low retention rate could be enhanced if the parents have college education (Hoyt, 1999). Accordingly, for fall-to-fall semester student retention if the parents have some education the level would be higher, while when it comes to fall-to-spring semester the fathers having college education is a predictor of good student retention, while the mothers having a college education has a negative impact. Nevertheless, there is a need to do more research to find out why the findings are as they are. According to Lee, Sax, and Hagedorn (2004) parents with post-secondary education could be in the know about what takes place in colleges and could tell their children about the commitment that is necessary to succeed in school work, including what kind of time to spend on their studies, what kind of financial obligations to expect, textbooks required, and generally the amount of sacrifice required in order to complete their education. Based on that Kinzie (2007) had stated that students who talk with their parents frequently and follow their guidance will not only be in a better position to complete their education, but in their stay in school their participation level would also be above average. The number of hours committed are also good indicators of retention, while at the same time the number of semester hours dropped in the first fall semester increases the odds of lower retention. Gender and ethnicity are not significant predictors of retention.
What Awaits the Future Research in Students Retention
Various sources such as Boylan (2002) and Adelman (1999) had indicated that developmental education deserves better consideration in most high learning establishments since some of them do not even recognize its existence. Furthermore, Redden (2007) had indicated the practice of universities that outsource remedial education to community colleges showing that they would avoid engaging in them directly. The reason for that might be students allowed in the universities should meet a higher-grade requirement that will eliminate the need for such programs. On another context, there is a pervasive growth on the prevalence of online courses that cannot be disputed because of the benefit outlined, however, its effectiveness is still shrouded and it requires some clarification. In spite of the ongoing debate about the quality and effectiveness of online education, its sheer existence had proven to be a predictor to student retention. This is due to the fact that there is enough evidence that indicates there are many students who quit their education because of the inflexible schedule that becomes intolerable burden in what they do such as their professional lives. Because of that, there is a call for more qualitative online courses. When it comes to financial aid, the reality had always been that it adds to the students persistence, although why it is so is not clear and it is among areas that require more research.
As far as limitations are concerned, there were some vital data missing in areas such as parents education, although the outcome was not affected in any significant way. Furthermore, it was not different from the other factors that were self-reported such as ethnicity, age, and gender that had not merited from verification.
Conclusion
Many others, including Lau (2003) believe that student retention is a crucial issue and there is a need to come up with effective methods to implement it properly, because there is a need to keep qualified students at higher learning establishments. Any failure in this respect will result in letting down students who otherwise would have had the opportunity to realize their dream of furthering their education. There will also be an impact on the system in the long run if the rate of the dropouts continue to rise since there will be a shortage of qualified personnel. The other adverse outcome could be seen from what a household of high school graduates make, which is at the range of 44,000 a year, whereas a household with a bachelor degree can make up to 80,300 and a household run by a graduate degree holder can earn up to 104,294 showing that education is vital for a high earning potential that has a repercussion that could be felt for a long time to come since children will also be affected by the earning capacity of their parents.
Another area believed to be crucial is the availability of data and its effective compilation, because it avails insight into the variables that have the most influence on the student retention level. Such data could also be employed to come up with effective methods that will facilitate the understanding of the retention problem better so that it will be possible to introduce effective interventions that will enhance the persistence level of the students. All the discussed factors and variables had been found through this research to augment the persistence level of the students, while at the same time they are important predictors of retention.
APPENDIX
HYPERLINK fileCDocuments20and20SettingsUserMy20DocumentsPredictors20of20First-Year20Student20Retention20in20the20Community20College.htm l toctoc o Table 1 Fall 2001-2004 First-Time-In-College Student Descriptive Statistics (N 9,200) Table 1 Fall 2001-2004 First-Time-In-College Student Descriptive Statistics (N 9,200)
Legend for Chart
A - Variable
B - Explanation
C - N
D -
E - Median
F - M
G - SD
A
B C D
E F G
StudentSexM
Gender (male) 4,030 43.8
StudentSexF
Gender (female) 5,163 56.1
StudentSexUnknown
Gender (unknown) 7 0.1
StudentStartAge
Age at college entrance
19 23.58 8.64
StudentEthnicHispa
Ethnicity (Hispanic) 2,155 23.4
StudentEthnicWhite
Ethnicity (White) 6,113 66.4
StudentEthnicOther
Ethnicity (Other) 681 7.4
StudentEthnicUnknown
Ethnicity (unknown) 251 2.7
TookDevMath
Took developmental mathematics class 6,009 65.3
PassedDevMath
Passed developmental mathematics class 3,036 50.5
TookDevWriting
Took developmental writing class 514 5.6
PassedDevWriting
Passed developmental writing class 280 54.5
TookDevReading
Took developmental reading class 2,042 22.2
PassedDevReading
Passed developmental reading class 1,280 62.7
EnrolledInSSS
Enrolled in TRIO Student Support Services 357 3.9
RecvFinancialAid
Received financial aid 5,539 60.2
EduMotherSomeCollege
Mothers educational attainment (some college) 2,569 27.9
EduFatherSomeCollege
Fathers educational attainment (some college) 2,264 24.6
TookInternet
Took an Internet class 3,103 33.7
HrsEnrolled1stFall
Hours enrolled during the first fall semester
12 9.97 3.92
HrsDropped1stFall
Hours dropped during the first fall semester
0 1.82 3.23
EnrolledNextTerm (Fall-Spring)
Enrolled in the subsequent spring term 6,324 68.7
EnrolledNextYear (Fall-Fall)
Enrolled in the subsequent fall term 4,430 48.2
HYPERLINK fileCDocuments20and20SettingsUserMy20DocumentsPredictors20of20First-Year20Student20Retention20in20the20Community20College.htm l toctoc o Table 2 Correlations (r) of Retention With Predictors of Retention for First-Time-In-College Students, Fall 2001-2004, by Year (N 9,200) Table 2 Correlations (r) of Retention With Predictors of Retention for First-Time-In-College Students, Fall 2001-2004, by Year (N 9,200)
Legend for Chart
A - Fall 2001 Next Term
B - Fall 2001 Next Year
C - Fall 2002 Next Term
D - Fall 2002 Next Year
E - Fall 2003 Next Term
F - Fall 2003 Next Year
G - Fall 2004 Next Term
H - Fall 2004 Next Year
A B C
D E F G H
StudentSexM -.054 -.036 -.018
-.036 -.057 -.084 -.032 -.075
StudentStartAge -.116 -.101 -.075
-.117 -.083 -.099 -.033 -.101
StudentEthnicHispanic -.022 -.029 -.014
-.025 .026 -.017 -.022 .004
StudentEthnic White .026 .041 .016
.040 -.017 .018 .029 .039
StudentEthnicOther -.011 -.025 -.005
-.031 -.013 -.004 -.015 -.073
PassedDevMath .274 .243 .214
.253 .225 .257 .247 .235
PassedDevWriting .259 .126 .183
.377 .203 .301 .365 .455
PassedDevReading .494 .387 .353
.388 .386 .399 .422 .427
EnrolledlnSSS .100 .111 .090
.122 .092 .115 .098 .086
RecvFinancialAid .246 .220 .262
.216 .206 .199 .219 .154
EduMotherSomeCollege .022 .031 -.018
.022 -.005 .016 -.011 -.005
EduFatherSomeCollege .037 -.011 -.001
-.007 .071 .083 .066 .060
TookInternet .242 .299 .243
.325 .239 .330 .224 .297
HrsEnrolled1stFall .246 .179 .280
.199 .265 .173 .262 .160
HrsDropped1stFall -.218 -.155 -.177
-.142 -.210 -.166 -.191 -.133
p .05 (two-tailed). p .01 (two-tailed).
HYPERLINK fileCDocuments20and20SettingsUserMy20DocumentsPredictors20of20First-Year20Student20Retention20in20the20Community20College.htm l toctoc o Table 3 Correlations of Retention and Predictors of Retention for FTIC Students, 2001-2004, With All Years Combined Table 3 Correlations of Retention and Predictors of Retention for FTIC Students, 2001-2004, With All Years Combined
Legend for Chart
A - Enrolled Next Term (Fall to Spring)
B - Enrolled Next Year (Fall to Fall)
A
B
StudentSexM r -.040, p .001, n 9,193
r -.058, p .001, n 9,193
StudentStartAge r -.077, p .001, n 9,196
r -.104, p .001, n 9,196
StudentEthnicHispa r -.007, p .511, n 8,949
r -.017, p .109, n 8,949
StudentEthnicWhite r .013, p .226, n 8,949
r .035, p .001, n 8,949
StudentEthnicOther r -.011, p .287, n 8,949
r -.034, p .001, n 8,949
PassedDevMath r .241, p .001, n 6,009
r .248, p .001, n 6,009
PassedDevWriting r .262, p .001, n 514
r .358, p .001, n 514
PassedDevReading r .409, p .001, n 2,042
r .403, p .001, n 2,042
EnrolledlnSSS r .094, p .001, n 9,200
r .108, p .001, n 9,200
RecvFinancialAid r .233, p .001, n 9,200
r .197, p .001, n 9,200
EduMotherSomeCollege r .001, p .956, n 5,339
r .024, p .083, n 5,339
EduFatherSomeCollege r .037, p .008, n 5,224
r .025, p .068, n 5,224
TookInternet r .238, p .001, n 9,200
r .312, p .001, n 9,200
HrsEnrolled1stFall r .264, p .001, n 9,200
r .178, p .001, n 9,200
HrsDropped1stFall r -.199, p .001, n 9,200
r -.149, p .001, n 9,200
HYPERLINK fileCDocuments20and20SettingsUserMy20DocumentsPredictors20of20First-Year20Student20Retention20in20the20Community20College.htm l toctoc o Table 4 Logistic Regression Model Predicting Retention Fall to Spring (N 9,196) Table 4 Logistic Regression Model Predicting Retention Fall to Spring (N 9,196)
Legend for Chart
A - B
B - SE
C - Wald
D - df
E - p
F - Exp(B)
G - 95 CI for Exp(B) Lower
H - 95 CI for Exp(B) Upper
A B C D E
F G H
StudentStartAge 0.011 .003 12.223 1 .001
1.011 1.005 1.018
EnrolledInSSS 0.803 .198 16.439 1 .001
2.232 1.514 3.291
RecvFinancialAid 0.473 .054 75.631 1 .001
1.605 1.443 1.786
TookInternet 0.947 .062 233.816 1 .001
2.577 2.282 2.909
HrsEnrolled1stFall 0.153 .008 408.455 1 .001
1.165 1.148 1.182
HrsDropped1stFall -0.156 .008 386.737 1 .001
0.856 0.843 0.869
PassedDevMath 0.762 .073 110.121 1 .001
2.143 1.858 2.470
NoDevMath -0.245 .065 14.010 1 .001
0.783 0.688 0.890
PassedDevReading 1.197 .118 103.647 1 .001
3.310 2.629 4.168
NoDevReading 0.787 .089 78.179 1 .001
2.197 1.845 2.616
EduMotherSomeCollege -0.157 .070 5.077 1 .024
0.855 0.745 0.980
EduFatherSomeCollege 0.247 .073 11.433 1 .001
1.280 1.110 1.478
Constant -2.064 .145 203.590 1 .001
0.127
Note CI Confidence Interval.
HYPERLINK fileCDocuments20and20SettingsUserMy20DocumentsPredictors20of20First-Year20Student20Retention20in20the20Community20College.htm l toctoc o Table 5 Logistic Regression Model Predicting Retention First Fall to Second Fall (N 9,200) Table 5 Logistic Regression Model Predicting Retention First Fall to Second Fall (N 9,200)
Legend for Chart
A - B
B - SE
C - Wald
D - df
E - p
F - Exp(B)
G - 95 CI for Exp(B) Lower
H - 95 CI for Exp(B) Upper
A B C D E
F G H
EnrolledInSSS 0.756 .138 30.109 1 .001
2.129 1.625 2.789
RecvFinancialAid 0.342 .051 44.642 1 .001
1.408 1.274 1.557
TookInternet 1.151 .052 499.569 1 .001
3.163 2.859 3.499
HrsEnrolled1stFall 0.067 .007 104.403 1 .001
1.069 1.056 1.083
HrsDropped1stFall -0.111 .008 185.373 1 .001
0.895 0.881 0.909
PassedDevMath 0.698 .061 132.515 1 .001
2.011 1.785 2.264
NoDevMath -0.412 .061 45.994 1 .001
0.662 0.588 0.746
PassedDevWriting 0.704 .214 10.866 1 .001
2.023 1.331 3.075
NoDevWriting 0.090 .158 0.327 1 .567
1.095 0.803 1.492
PassedDevReading 1.184 .116 104.846 1 .001
3.267 2.605 4.098
NoDevReading 0.978 .100 95.436 1 .001
2.660 2.186 3.236
EduMotherSomeCollege 0.137 .063 4.771 1 .029
1.147 1.014 1.297
EduFatherSomeCollege 0.184 .065 7.910 1 .005
1.202 1.057 1.366
Constant -2.384 .186 164.435 1 .001
0.092
Note CI Confidence Interval.
Source Adapted from
2008 North Carolina State University
Categories:
0 comments:
Post a Comment