Factors shaping the probability of community vs. four-year college entrance and acquisition of the B.A. degree.
Community college proponents emphasize the open door nature of these institutions. Community colleges, they argue, have democratized higher education by providing an opportunity for all to enter the higher educational system. They fail to note, though, that the vast majority of public four-year institutions were virtually open-door prior to 1960.(2) Instead of democratizing higher education, critics suggest that community colleges effectively divert students, especially those from modest social origins, away from the four-year institutions they would have entered if community colleges were not an alternative.(3)
Thus, community college critics(4) see the community college reinforcing inequalities in the higher educational system that exist in the world at large. Today, community colleges are the point of entry for a majority of students.(5) Since 1978, women have outnumbered men in the community college population. Over half (53%) of female undergraduates were enrolled in public community colleges in 1988, compared with 48% of male undergraduates.(6)
African American students are increasingly concentrated in community colleges. Between 1976 and 1982, the number of African Americans enrolled in community colleges increased by 60,200, in comparison to an increase of only 8,300 at four year] institutions. By 1986, 47% of African Americans, who were enrolled in college, were enrolled in community colleges (up 5% from 1980).(7)
Karabel's early work shows individuals from lower socioeconomic backgrounds are more likely to attend community colleges than others. Community college entrants come from lower social class backgrounds as measured by parental income, occupation, and education than four-year college entrants.(8) Nineteen consecutive surveys of college freshmen, by the American Council of Education, have shown that the social composition of the community college population is lower than at four-year institutions. In 1984, the proportion of students from families with incomes less than $25,000 at community colleges, four-year colleges, and universities was 48.3%, 39.9%, and 28.7% respectively.(9) Two-year colleges, therefore, function as gates of entry into the higher educational system for large numbers of less economically advantaged students as well as for women and minority students.
This work analyzes factors that shape the likelihood of community college entrance compared to four-year college entrance with background variables constant. In modeling type of first college entered, it is important to control for the effects of SES, IQ, race, sex, and whether or not a respondent has as a goal completion of the Bachelor's degree. It is estimated that thirty to forty percent of all community college entrants aspire to achieve the B.A. degree.(10) In addition, the effect of region will be examined. Community colleges are more prevalent outside the South, thus respondents residing in Southern states should be less likely to enter community colleges than four-year institutions because there are fewer community college institutions in the area.
In their work, Alexander, Holupka and Pallas(11) find that the probability of four year college attendance is higher for blacks than others. Their model fails to control for the effect of region. Given the small number of traditionally black four-year colleges in the North, I will also include an interaction term in the preferred model (race X region). This allows for an examination of differences in likelihood of community college entrance for blacks vs. whites living in the North. I expect that blacks will be more apt to enter community colleges outside the South because relatively high numbers of black college entrants in the South will enter historically black four-year colleges. For example, half of all black four-year public college students in the state of Virginia attend Norfolk State University or Virginia State University-the two public black four-year institutions in the state.(12) Entrance requirements at these institutions mirror those at community colleges.
There is evidence that going to college is still perceived by parents (and students) as more appropriate for men than women.(13) We know little, though, about the likelihood of entering college, especially a community or four-year college, and how this varies by gender and ability. Thus an interaction term, sex X IQ, will also be included in the preferred model. I expect that women, even women who perform well on IQ tests, will be more likely to enter a community college than a four-year institution vs. comparable men.
Next, I model the likelihood of achieving the Bachelor's degree. The dependent variable will be completion of the Bachelor's (B.A./B.S.) degree or not. This model will be the basic one for community college entrance with the addition of a variable for type of first college entered.
It is well known that community college entrance negatively impacts the chances of a B.A. aspirant completing the degree compared to four-year college entrants. Three national surveys show that, on average, the vast majority (70%) of four-year college entrants had received the B.A. degree (within four years), while approximately a fourth (26%) of public community college entrants had achieved the degree.(14)
In their work, Alba and Lavin(15) found that students who entered a four year college in the City University of New York system (CUNY) had a distinct advantage in likelihood of B.A. completion compared to comparable CUNY community college entrants. Using regression (OLS) analysis, they found community college entrance had a modest negative effect on successfully completing one's educational aspirations net of background variables.
Cohen and Brawer(16) argue that the baccalaureate gap exists because of selection bias differences between students who enter community colleges instead of four year schools. Thus, in modeling B.A. attainment it is critical to control for the effects of socioeconomic background, ability, educational goals, and race. Mine is a conservative test of a link between type of college entered and degree success as I do not restrict completion of the degree to a four year time period.
THE B.A. ATTAINMENT GAP
It is well known that community college entrants tend to be drawn from less advantaged SES backgrounds than four-year entrants. Entering a community college may well be perceived, by both parents and students coming from a less advantaged SES position than others, as an "appropriate" college choice. For students coming from lower parental SES backgrounds, the idea of acquiring the four-year college Bachelor degree may be one mixed with conflict. The lower the SES ranking, the less education, in general, the respondent's father, mother, and oldest older sibling acquired. An independent negative association between SES and acquisition of the B.A. degree bodes ill for a community college entrant's degree success as a disproportionate number of community college entrants are drawn from low SES backgrounds.
Some see the community college as a place where students can demonstrate to others that they are "college material". A problem with this view is an assumption that the two playing fields are fair and equal. As Dougherty(17) writes
The poorer outcomes encountered by baccalaureate aspirants entering
community colleges compared to four-year colleges cannot be attributed
solely to the community college entrants' less advantaged backgrounds,
poorer academic preparation, and lower aspirations. When we compare
students with similar traits, we find that baccalaureate aspirants are still
significantly less likely to realize their hopes if they enter a community
college than a four-year college. There is a quite sizable institutional
effect, over and above the effect of differences in student characteristics.
To explore this idea, an interaction term, IQ X type of first college entered, will be included in the preferred model. I expect that IQ will be more important in shaping the likelihood of degree completion for community college than four-year college entrants. Community college entrants must prove themselves in a way that is not true for four-year entrants.
To add empirical support to the idea that an institutional effect exists in shaping B.A./B.S. completion between community and four-year college entrants, I will include a second interaction term in the preferred model (goal X type of first college entered). I expect that it will be much less likely that a community college entrant, who wants to finish the Bachelor's degree, acquires the degree than a comparable four-year entrant. Two and four-year institutions are vastly different kinds of institutions. Community college entrance, in itself, presents a barrier to degree completion for community college entrants. Estimates are that only 15% of community college entrants eventually transfer to a four-year school.(18) Community college students, even those aspiring to acquire the B.A. degree, receive less advice, less encouragement, and less systematic academic preparation than four-year college entrants.(19)
DATA SET AND SAMPLE SELECTION
The sample was selected from the National Longitudinal Survey of Labor Market Experiences data for young men and young women. Each cohort contains approximately 5,000 respondents aged 14-24 in the base year data collection began. The base year for men was 1966 and for women 1968. My sample includes all men and women who entered a community or four year college between the years 1966 and 1978. I included only those students with full data on educational status and IQ. Of the 1,796 respondents in my restricted sample, 29% entered a community college.
The dichotomous nature of the dependent variables (community college entrant or not; Bachelor's recepient or not) violates assumptions of ordinary least squares regression analysis. Therefore, logistic analysis was chosen because it provides the power of regression analysis and yields parameters that enable the estimation of the effects of independent variables on the odds of a dichotomous dependent variable.(20) In this model, the regression parameters indicate a unit change in the log of the odds of community college entrance or acquisition of the Bachelor's degree. The additive coefficients (b) were transformed to their multiplicative counterparts [e.sup.b]. Focusing on multiplicative effects eases interpretation by removing the logarithm from both sides of the equation. This allows one to analyze the effects of a unit change in the independent variables on the odds of falling into category one of the dependent variable (which in my study is entering a community college or not; acquistion of the Bachelor's or not). An odds ratio [e.sup.b] greater than 1.0 indicates an increased likelihood of community college entrance/ B.A. acquisition, while an odds ratio less than 1.0 indicates a decreased likelihood of the event occurring. Thus the direction of an odds ratio, whether it is greater or less than 1.0, "can be thought of as its `sign'."(21)
SES: Is an equally weighted sum of father's education, mother's education, father's occupational status when respondent was 14, educational attainment of respondent's oldest older sibling, and a measure of availability of reading material in the home when respondent was 14. See Table 1 for the means and standard deviations of independent variables in the model.
Table 1. Means and Standard Deviations of Independent Variables Vanables Mean s.d. Race .88 .32 IQ .36 .92 SES .55 .87 Region .68 .46 Sex .53 .50 B.A. is Goal .84 .37 Type of First College Entered .71 .46
Mental ability (IQ) test scores were obtained from the respondent's last high school. These scores were normalized.
Region is a dummy variable that distinguishes respondents who live outside the South from all others. Southern states include: Delaware, Maryland, Washington, DC, Virginia, West Virginia, North Carolina, South Carolina, Georgia, Florida, Kentucky, Tennessee, Alabama, Mississippi, Arkansas, Louisiana, Oklahoma, and Texas.
Race was dichotomized, and coded 1 for white and 0 for black respondents. The majority of black respondents (71%) entered a four-year college.
Sex is coded 1 if male, 0 if female.
Bachelor's degree is goal is a dummy variable that distinguishes respondents who want to obtain at least a Bachelor's degree from those who do not have this as a goal.
Type of first college entered is a dummy variable and coded 1 if respondent was a four year entrant and 0 if a community college entrant.
As expected, students coming from a more advantaged parental socioeconomic background (.80) and those who score higher on tests of mental ability (.72) have lower odds of entering a community college than a four-year institution (see Table 2). The odds of men entering a community college rather than a four-year school are greater than for women by a factor of .52. Not surprisingly, students who have the Bachelor's as a goal are less likely (.27) to enter a community college than a four-year institution. The odds of students entering a four-year school not a community college, who have the Bachelor's degree as an educational goal, are nearly four times the odds of those who do not have this as a goal. The coefficient for race did not reach statistical significance in this model.
Table 2. Effects of Predictor Variables on the Likelihood of Entering a Community College-Basic and Preferred Models Basic Model Preferred Model Variables b e(b) s.e. b e(b) s.e. SES -.22 .80 .07 -.23 .79 .07 IQ -.33 .72 .06 -.24 .79 .08 Region .38 1.46 .12 1.62 5.05 .35 Race .35 1.41 .19 1.22 3.39 .32 Sex .42 1.52 .11 .51 1.66 .12 B.N is Goal -1.30 .27 .14 -1.28 .28 .14 IQ X Sex -.30 .74 .12 Race X Region -1.42 .24 .37 Model [x.sup.2] = 179.41 df = 6 p = 0 (Basic Model) Model [x.sup.2] = 198.24 df = 8 p =.0 (Preferred Model) All variables in basic model significant at .05 level except Race which is significant at .10 level. All variables in preferred model, significant at .05 level.
The effect of the variable region on type of first college entered was also explored. As hypothesized, living outside the South (1.46) is positively associated with community college entrance. The interaction term, race X region, shows that if you live outside the South then the odds are higher that you will go to a community college and that this effect is much stronger for blacks (5.05) than whites (1.22). These findings do not support the work of Alexander and his colleagues (22) who posit that blacks are more likely, on the average, than whites to enter a four-year college. It appears region is an important variable in modeling variation in type of college entered probabilities by race. The advantaged enrollment prospects of black students, that Alexander et al. pick up, is in reality only an advantage for blacks in the South given enrollment in historically black public four-year colleges. Again enrollment qualifications at these schools tend to be more in line with those at community colleges rather than other four-year institutions.
The other interaction term included in the preferred model, IQ X sex, indicates that with each unit increase in IQ the odds of entering a community college decrease. This effect is more pronounced for men (.58) than women (.79). This finding lends support to my hypothesis that women who perform well on IQ tests are more likely to enter a community college than a four-year school vs. comparable men. In other words, "being smart" does more to boost the odds of four-year college entrance for men, on average, than for women.
ACQUISITION OF THE B.A. DEGREE
As expected, four-year college entrance is positively associated with completion of the B.A. degree all else equal. The odds of four-year college entrants acquiring the Bachelor's degree are greater than those of community college entrants by a factor of .36.
As hypothesized, students coming from advantaged parental SES backgrounds had higher odds of finishing the Bachelor's degree. With each unit increase in SES, the odds of completing the Bachelor's degree increase by a factor of .21. In other words, the more formal education the respondent's parents and oldest sibling acquired boosts her/his odds of finishing the Bachelor's degree all else equal. Community college entrants tend to come from less advantaged parental SES backgrounds, thus this association does not bode well for their successfully completing the degree.
Besides four-year college entrance and coming from a higher SES background, having the Bachelor's as a goal, being male, and having higher measured mental ability are also positively associated with acquisition of the Bachelor's degree. Variation in race and region have virtually no independent effect on finishing the degree when controlling for other background variables.
As hypothesized, increased mental ability significantly bettered the odds of finishing the Bachelor's degree for community college entrants more so than for four-year entrants. With each standard deviation increase in IQ, the odds of getting the degree are higher for community college (1.72) than four-year college (1.14) entrants all else equal. Community colleges entrants, who go on to acquire the B.A. degree, are the most able students (as measured by IQ) in the community college population. This finding supports the work of Lee and Frank(23) who write
Since it is the students who are relatively "better" on entry who come out of
community colleges with academic preparation that is strong enough to enable
them to continue their education, we must conclude that these institutions are
not reducing social stratification in academic higher education to any great
The second interaction term included in this model, goal X type of first college entered, shows that the odds of finishing the Bachelor's degree, if this is one's goal, are significantly higher if one first enters a four-year rather than a community college. The odds of getting a B.A. degree for four-year entrants, who have this as an educational goal, are more than three times higher than for comparable community college entrants. It appears that community colleges (and perhaps four year institutions as well) are not facilitating transfer for students who wish to pursue the B.A. degree.
One could argue that since not all community college entrants want to acquire the Bachelor's degree, the model should be restricted to respondents who have B.A. completion as an educational goal. To address this concern, the sample was restricted to those who had as a goal acquisition of the degree (84% of my sample wanted to acquire the Bachelor's degree). I deleted goal as a predictor variable in this model. The results of this work show that the odds of four-year college entrants are 1.63 times the odds of community college entrants of acquiring the degree (see Table 3 Column 3). Only socioeconomic background and ability are better predictors of success in finishing the Bachelor's degree than community college entrance.
Table 3. Effects of Predictor Variables on the Likelihood of Acquiring the Bachelor's Degree-Basic and Preferred Models Basic Model Basic Model Preferred Model (Restricted Model) Variables b e(b) s.e. b e(b) s.e. b e(b) s.e. SES .19 1.21 .06 .19 1.21 .06 .22 1.25 .07 IQ .22 1.24 .06 .54 1.72 .13 .24 1.28 .06 Region .02 1.02 .11 .03 1.03 .11 -.04 .96 .12 Race .04 1.04 .17 .05 1.05 .17 .02 1.02 .19 Sex -.17 .84 .10 -.17 .84 .10 -.17 .84 .11 B.A.is Goal 1.39 3.97 .15 .84 2.32 .20 Type of FCE .31 1.36 .11 -.49 .61 .27 .49 1.63 .12 IQ X Type -.41 .66 .15 Goal X Type 1.10 3.00 .30 Model [X.sup.2] = 183.84 df = 7 p = .0 (Basic Model) N = 1796. Model [X.sup.2] = 203.73 df = 9 p = .0 (Preferred Model) Model [X.sup.2] = 60.15 df = 6 p = .0 (Basic restricted Model) N = 1510. All variables in basic model (and basic restricted model) are significant at the .05 level except region and race. All variables in the preferred model significant at .05 level.
Compared to four year college entrance, community college entrance exerts a deliterious effect on the odds of achieving a Bachelor's degree all else equal. This, coupled with the fact that students coming from a less advantaged parental socioeconomic background are more likely to enter a community college not a four year school, raises equality of opportunity questions. Entering a community college introduces a structural barrier to completion of the Bachelor's degree. Regardless of ability, motivation to acquire the degree, gender, or race, the mere act of beginning one's higher education at a community college makes it less likely that a student will ever complete the degree. Channeling disproportionate numbers of low SES students into community colleges guarantees a double negative status in their quest for a Bachelor's degree. Even for community college entrants who want to acquire the degree and for those who score well on IQ tests, community college entrance has a deleterious effect on degree completion.
(1) Steven Brint and Jerome Karabel, The Diverted Dream (Oxford: Oxford University Press, 1989). (2) Richard Ferrin, A Decade of Change in Free-Access Higher Education (New York: College Entrance Examination Board, 1971). (3) Steven Brint and Jerome Karabel, The Diverted Dream (Oxford: Oxford University Press, 1989). (4) Fred Pincus, "The False Promises of Community Colleges: Class Conflict and Vocational Education," Harvard Educational Review, 50(1980):333-361. (5) U.S. Department of Education, National Center for Education Statistics: Digest of Educational Statistics (Washington, DC: Government Printing Office, 1986). (6) U.S. Department of Education, National Center for Education Statistics: Digest of Educational Statistics (Washington, DC: Government Printing Office, 1990). (7) Linda Darling-Hammond, Equality and Excellence (New York: The College Board, 1985). (8) Leland Medsker and J.W. Trent, The Influence of Different Types of Public Higher Institutions on College Attendance from Varying Socioeconomic and Ability Levels (Berkeley: Center for Research and Development in Higher Education, 1965). (9) Jerome Karabel, "Community Colleges and Social Stratification," Harvard Educational Review, 42(1972):521-562. (10) Kevin Dougherty, "The Effects of Community Colleges: Aid or Hindrance to Socioeconomic Attainment?," Sociology of Education, 60(1987):86-103. (11) Karl Alexander, Scott Holupka and Aaron Pallas, "Social Background and Academic Determinants of Two-Year versus Four-Year College Attendance: Evidence from Two Cohorts a Decade Apart," American Journal of Education, 96(1987):56-80. (12) Fact Book 1988-89 of Higher Education in Virginia. Virginia (1988-1989), pp. 17-21, 22-28. (13) Mirra Komarovsky, Women in College (New York: Basic Books, Inc., 1985). (14) William Valez, "Finishing College: The Effects of College Type," Sociology of Education, 58(1985):191-200. (15) Richard Alba and David Lavin, "Community Colleges and Tracking in Higher Education," Sociology of Education, 54(1981):223-237. (16) A.M. Cohen and F. Brawer, The Collegiate Function of Community Colleges (San Francisco: Jossey-Bass, 1987). (17) Kevin Dougherty, "The Community College at the Crossroads," Harvard Educational Review, 61(1991):311-336. (18) Fred Pincus and E. Archer, Bridges to Opportunity? (New York: The College Board, 1989). (19) Kevin Dougherty, "Community College and Baccalaureate Attainment," Journal of Higher Education, 63(1992):188-214. (20) John Aldrich and Forrest Nelson, Linear Probabilities, Logit, and Probit Models (Beverly Hills, CA: Sage, 1984). (21) Philip Morgan and Jay Teachman, "Logistic Regression: Description, Examples, and Comparisons," Journal of Marriage and the Family, 50(1988): 929-936. (22) See Karl Alexander, Scott Holupka and Aaron Pallas, "Social Background and Academic Determinants of Two-Year versus Four-Year College Attendance," American Journal of Education, 96(1987):56-80. (23) See Valerie Lee and Kenneth Frank, "Students' Characteristics that Facilitate the Transfer from Two-Year to Four-Year."
ELIZABETH MONK-TURNER, Direct all correspondence to: Elizabeth Monk-Turner, Department of Sociology, Old Dominion University, Norfolk, Virginia 23529 0090. Telephone: (804) 683-3801.
|Printer friendly Cite/link Email Feedback|
|Publication:||The Social Science Journal|
|Date:||Jul 1, 1995|
|Previous Article:||The rise, hegemony, and decline of the Chicago School of Sociology, 1892-1945.|
|Next Article:||Effects on the AFDC-Basic caseload of providing welfare to two-parent families.|