Search

In this study, I draw on Bahr's ("Research in Higher Education" 51:724-749, 2010; New Directions for Institutional Research S1:33-48, 2011) behavioral typology of first-time community college students to examine college-level variation in students' patterns of use of 105 community colleges in California. I find that students' patterns of use vary greatly across the colleges, and, further, these patterns tend to cluster in such a fashion that colleges may be classified based on dominant or disproportionate patterns of use. Using "k"-means cluster analysis, I identify five types of community colleges, including "Community Education Intensive," "Transfer Intensive," "Workforce Development Intensive," "High-Risk Intensive," and "Mixed Use." I describe each of these community college types and then investigate whether the patterns of student use that characterize the identified types appear to be primarily a consequence of institutional policies and practices or, conversely, a product of localized community demand and the associated circumstances and choices of the students who attend a given college. The evidence, though limited, tends to support the latter: variation in patterns of student use across institutions appears to be primarily a product of localized community demand. Finally, I draw on established performance indicators to examine the implications of the identified patterns of student use for observed institutional performance. I find that institutions that differ in terms of dominant or disproportionate patterns of student use also differ significantly and systematically on a number of measures of institutional performance.

Two related themes currently dominate discourse on open-access colleges, particularly community colleges: increasing college-going and degree attainment and improving the performance of postsecondary institutions with respect to producing graduates. Largely missing from this discourse, however, is cogency concerning the innumerable ways in which students use open-access institutions, and the ways in which students' patterns of use interact with institutional policies and practices to influence the outcomes that they experience. Absent a thorough understanding of students' pathways through the institution, the development of interventions and the adjustment of institutional policies and practices to improve students' outcomes will be more a product of guesswork than of sound empirical reasoning. Unfortunately, traditionally favored analytical approaches are unlikely to rectify this large and troubling gap in our understanding. In this essay, I present the case for a different approach--a "deconstructive" approach--to illuminate students' pathways and the relationships between these pathways and outcomes. (Contains 3 notes.)

Nationally, a majority of community college students require remedial assistance with mathematics, but comparatively few students who begin the remedial math sequence ultimately complete it and achieve college-level math competency. The academic outcomes of students who begin the sequence but do not complete it are disproportionately unfavorable: most students depart from the community college without a credential and without transferring to a four-year institution. Interestingly, however, many of these students continue to attend the community college after they exit the remedial math sequence, sometimes for an extended period. One is led to ask why students who do not complete the sequence generally are not finding their way to an alternative credential objective that does not require college-level math competency, such as a career and technical education certificate, sometimes referred to as a vocational certificate. In this study, I explore three possible answers to this question, including difficulty navigating to the alternative credential, declining participation in the community college, and declining academic performance. I find that all three of these explanations contribute (to varying degrees) to explaining the low rate of certificate completion among remedial math students who do not achieve college-level math competency.

Each year, a sizeable percentage of community college students enroll in remedial coursework to address skill deficiencies in math, writing, and/or reading. Unfortunately, the majority of these students do not attain college-level competency in the subjects in which they require remedial assistance. Moreover, students whose point of entry into the remedial sequence is at the lower end of the hierarchy of skill suffer the lowest rates of attainment by far. Yet, to date, we do not understand fully why students who begin at the lower end of the remedial sequence are so much less likely than are students who begin at the higher end to attain college-level competency. The purpose of this study is to illuminate the junctures in the remedial sequences in math and writing at which meaningful attrition of students is occurring and, in particular, the junctures at which "low-skill" remedial students suffer differential attrition relative to "high-skill" remedial students. To accomplish this end, I use data that address students in California's community colleges to examine three ways of characterizing and explaining the differential in college-level skill attainment between low- and high-skill remedial math students and, separately, between low- and high-skill remedial writing students. The three characterizations include nonspecific attrition, skill-specific attrition, and course-specific attrition. I find that each of these characterizations contributes to explaining the differential in college-level skill attainment between low- and high-skill remedial students.

The traditional unidirectional ("linear") postsecondary path from high school to a community college to a 4-year institution into the workforce represents accurately a decreasing proportion of the pathways actually taken by students through higher education. Instead, students increasingly exhibit patterns of enrollment that take them through multiple postsecondary institutions, both within levels of the higher education system (e.g., multiple community colleges, multiple 4-year institutions) and across levels (e.g., movement back and forth between community colleges and 4-year institutions). These "swirling" patterns of enrollment are widely recognized by scholars of higher education, but they remain poorly understood. In this study, I employ data that address 89,057 first-time students in the California community college system to answer a number of key questions concerning lateral transfer between community colleges, which, according to prior research, constitutes one sizeable component of student "swirl". Building on the very limited work on this topic, I examine whether the reported high prevalence of lateral transfer holds true under a more stringent operational framework than that employed in prior work. I explore whether lateral transfer is primarily an artifact of students enrolling simultaneously in multiple community colleges, sometimes called "double-dipping". I investigate the timing of lateral transfer from several different perspectives to determine how lateral transfer fits in students' progress and development. Finally, I probe the relationship between students' level of academic investment in their current community college and the risk of lateral transfer.

One useful and increasingly popular method of classifying students is known commonly as cluster analysis. The variety of techniques that comprise the cluster analytic family are intended to sort observations (for example, students) within a data set into subsets (clusters) that share similar characteristics and differ in meaningful ways from other subsets. In the case of community college students, for example, clusters may be formed on the basis of student aspirations, student course-taking and enrollment behaviors, student demographic characteristics, or any combination of these or other measures. These clusters assist in the identification of patterns of behaviors or characteristics in data sets that contain an otherwise incomprehensible amount of student information. In this article, the authors provide an introduction to the cluster analytic method as it pertains to research on community college students. The execution of cluster analysis requires a number of carefully considered methodological decisions. The first set of decisions concerns the selection of the variables that will be included in the analysis, the scaling of these variables, and the structure of the data. The second involves the selection of a proximity metric by which within-group homogeneity and between-group heterogeneity will be calculated. Then, the cluster technique itself must be selected. Finally, one must decide how to make sense of the identified clusters. The authors describe each stage of the decision-making process and detail several common approaches at each stage. They cite many comprehensive sources throughout to aid researchers in locating additional informational resources.

Students use the community college in a wide variety of ways to achieve an equally wide variety of ends. Some of these ends align closely with institutional goals, priorities, and performance indicators, and others do not. Consequently a typology of community college students based on their use of the institution has the potential to be of great informational and interpretive value to community college stakeholders. In this article, the author broadens and extends his previous work (Bahr, 2010a), presenting a revised classification scheme. He also demonstrates the utility of a behavioral classification scheme for interpreting findings of significance for policy and practice. In particular, he examines how a behavioral classification scheme may inform the interpretation of findings concerning one especially prominent issue in many community colleges today: postsecondary remediation. (Contains 4 tables.)

Recent research on mathematics remediation in community colleges indicates that there are large differences in the rate of successful remediation between students of differing levels of initial math deficiency. One might presume that these "skill gaps" in successful remediation are a result of differing rates of student retention. That is, presumably poorer math skills are associated with a shorter duration of college attendance (lower persistence) and, in turn, a shorter duration of college attendance is associated with a lower likelihood of successful remediation. However, the limited prior evidence does not appear to support this explanation for the "skill gaps" in successful remediation. In an effort to make sense of this incongruity, I executed a detailed analysis of the relationship between level of math deficiency at college entry, persistence, and successful remediation in math. In contrast to the conclusions of prior work, I find that differences in persistence contribute strongly to the "skill gaps" in successful remediation, although this contribution is masked by the fact that the relationship between persistence and the likelihood successful remediation is neither proportional nor equal across levels of initial math deficiency, both of which were implicit analytical assumptions of prior work on this topic. (Contains 3 tables, 4 figures and 2 footnotes.)

The development of a typology of community college students is a topic of long-standing and growing interest among educational researchers, policy-makers, administrators, and other stakeholders, but prior work on this topic has been limited in a number of important ways. In this paper, I develop a behavioral typology based on students' course-taking and other enrollment patterns during a seven-year observation period. Drawing on data for a population of 165,921 first-time college students, I identify six clusters of behaviors: "transfer," "vocational," "drop-in," "noncredit," "experimental," and "exploratory." I describe each of these student types in terms of distinguishing course-taking and enrollment behaviors, representation in the first-time student cohort, predominant demographic characteristics, and self-reported academic goal. I test the predictive validity of the classification scheme with respect to long-term academic outcomes. I investigate the relationships between the primary classification scheme and several alternative classification schemes. Finally, I demonstrate the replicability of the classification scheme with an alternate cohort of students. (Contains 6 footnotes and 7 tables.)

Postsecondary remediation is an important and highly contentious issue that has received comparatively little comprehensive attention with respect to evaluating its effectiveness. This study addresses two relatively unexplored aspects of remedial efficacy: the moderating effects of depth and breadth of underpreparation. I find that, regardless of the depth or breadth of students' initial deficiencies, students who achieve college-level competency in English and math generally exhibit similar levels of attainment. This finding supports the efficacy of remediation even for students who face the greatest deficiencies in English or math and those who have dual deficiencies in both English and math. (Contains 5 tables and 2 figures.)

In this study, I examine racial disparities in successful remediation in math. I first quantify a previously unidentified racial gap in successful remediation and then seek to explain this gap through a set of mediating and moderating variables. In addition, I test the relative efficacy of remediation across racial groups. (Contains 6 tables and 5 notes.)

Variables that address student enrollment patterns (e.g., persistence, enrollment inconsistency, completed credit hours, course credit load, course completion rate, procrastination) constitute a longstanding fixture of analytical strategies in educational research, particularly research that focuses on explaining variation in academic outcomes. However, nearly all measures of enrollment patterns are handicapped by untested assumptions about a more fundamental measure, namely students' rate of progress. In this paper, I first explain how a variety of widely used measures of enrollment patterns are inextricably linked to students' rate of progress. I then describe a method of modeling mathematically students' rate of progress that employs hierarchical (multilevel) discrete-time event history analysis of repeated events. I conclude with an empirical example of the application of this method in which I test several hypotheses concerning students' rate of progress through the remedial math sequence toward the outcome of college-level math competency. In addition to the utility of the method that is proposed here, the issues discussed in this paper have important practical implications for institutional research, particularly with respect to the use of the various measures of enrollment patterns to explain variation in students' attainment.

Lateral transfer (between community colleges) is second only to upward transfer (to a 4-year institution) among community college students' most common patterns of transfer. Yet, upward transfer is the focus of innumerable studies, while lateral transfer has received very little empirical attention. This study explores the occurrence and frequency of lateral transfer in California and its consequences for the measurement of one particular outcome, namely, completion of a credential. The results indicate that students transfer laterally quite frequently, leading to substantial undercounts in rates of credential completion when measured from the standpoint of a single community college or single district. Furthermore, the frequency of lateral transfer varies systematically with a number of student characteristics of recurrent interest in the literature, leading to exaggerated differences in the likelihood of credential completion between some groups of students and inaccurately attenuated differences between other groups, when measured under the single-college and single-district analytical frameworks. (Contains 8 tables and 1 figure.)

Burton Clark's proposition concerning the "cooling out" of underprepared students in community colleges has a controversial history and remains a point of contention. Central to Clark's description of the "cooling out" process is the academic counselor, whose job it is to dissuade underprepared students from goals perceived to be overambitious and ease these students into lesser, presumably better-fitting academic trajectories. In this study, I test a number of hypotheses concerning the effect of advising on students' chances of attaining their goals. I seek to determine what effect advising has on students' attainment, and whether this effect is dependent upon students' academic preparation, students' race/ethnicity, the racial/ethnic composition of the college, or the representation of underprepared students in the college. I use hierarchical discrete-time event history analysis to analyze data that address two subsets of the Fall 1995 cohort of first-time freshmen who enrolled in any of California's 107 semester-based community colleges. I find that advising is actively beneficial to students' chances of success, and all the more so for students who face academic deficiencies, which contradicts deductions drawn from Clark's description of the active role of counselors in the "cooling out" process.

Prior research has established that the depth and breadth of remedial need in basic skills (math and English) exhibited by a student at college entry are strongly and negatively associated with the likelihood of achieving college-level competency in those subjects (i.e., successful remediation). This well-documented finding is built upon a body of work employing either simple bivariate analyses or regression analyses that assume additive effects. Yet, there are reasons to suspect that multiple basic skill deficiencies, rather than exhibiting additive effects alone, may exhibit a negative multiplicative interaction effect on the likelihood of successful remediation. In this research, I test the hypothesis that the negative effect of math deficiency increases in magnitude with decreasing English competency. Although the data support this hypothesis, I find that this interaction does not have substantive importance in the face of the powerful direct effect of math deficiency on the likelihood of successful remediation in math.

Accountability in performance is a subject of increasing attention and concern for community colleges. In California, much attention has been focused on student transfers from the community colleges to four-year institutions as a primary facet of institutional performance. This focus often leads to comparisons of transfer rates between colleges with the goal of identifying low performing and high performing colleges. However, comparison of raw rates of transfer ignores important differences in the structural conditions and exogenous variables affecting the performance of each college, leading to inequitable stratification of colleges and inaccurate classification of colleges in terms of transfer performance. To treat this fundamental incomparability of raw transfer rates, a method for equitable comparison of transfer outcomes was developed and implemented in California's community college system. This method involves major enhancements to prior efforts in California to implement transfer comparisons, including a less biased definition of the transfer rate, the use of statistical models to adjust for exogenous variables observed to affect the transfer outcome, the use of multiple student cohorts, and the inclusion of data on student transfers to a wider range of four-year institutions (both private and public, and both in-state and out-of-state).

California's community college accountability strategy, the Partnership for Excellence program, includes, as an aspect of the outcome assessment component of the program, a mechanism to "level the playing field" between colleges. This function is accomplished through adjustment models, statistically derived equations that "adjust" for observed relationships between exogenous variables and college-level outcomes of interest. The development of adjustment models for each of the several outcomes has relied upon an exploratory process to derive a parsimonious set of exogenous variables with nonzero (statistically significant) relationships to the outcome of interest. One previously unmeasured adjustment variable has received considerable interest in discussions of the adjustment model development process, namely the academic preparedness of entering students at each college. This article addresses the work of the authors to develop a measure of student average academic preparation for use in "leveling the playing field" in community college outcome measurement and accountability.

This study investigates community college skills builders—students who enroll for a short time, take courses concentrated in career and technical education (CTE), are highly successful in their coursework, but typically leave college without a postsecondary credential. Drawing on administrative data from Colorado, Ohio, Michigan, and California, we use k-means cluster analysis to identify a distinct group of skills builders in each state. We analyze their prevalence, characteristics, course-taking, and educational and labor market outcomes. We then develop criteria that state systems or individual institutions can use to identify skills builders in their student populations. Skills builders comprised similar proportions of students in each state, between 11 and 14 percent, and were qualitatively similar across states. Their average earnings were flat or trending downward prior to college but reversed after college, rising at a statistically significant rate. Fields of study and certificate completion rates among skills builders varied across states. Identifying skills builders will strengthen state and institutional efforts to communicate student successes that cannot be measured with credential completions, while also equipping institutions with the information needed to better serve skills builders. [ABSTRACT FROM AUTHOR]