2015 - Course-Taking Patterns of Community College Students Beginning in STEM: Using Data Mining Techniques to Reveal Viable STEM Transfer Pathways

Attribution: Wang, Xueli
Researchers: Xueli Wang
University Affiliation: Wisconsin University
Email: xwang273@wisc.edu
Research Question:
This study examines the course-taking trajectories of beginning community college students, and the resulting transfer outcomes as related to STEM. The following question guides this research: What course-taking patterns are most contributive to upward transfer in STEM fields?
Published: Yes
Journal Name or Institutional Affiliation: Research in Higher Education
Journal Entry: Vol. 57, Pp. 544-569
Year: 2015
Findings:

– In general, ‘‘antecedent” course-taking patterns that result in transfer in STEM as a ‘‘consequent” involve a combination of ‘‘likely transferable” STEM courses and math courses in the earlier terms of students’ community college attendance. In particular, it is intriguing to note that, among STEM transfer students, despite the inevitable math-learning path, math course-taking during the very first term does not appear to be the most frequent course-taking pattern. Instead, the most viable course-taking trajectories contributing to STEM transfer, by and large, feature a pattern that first introduces ‘‘likely transferable” STEM courses during the first term, followed by math exposure during the subsequent terms.
– Non-transfer students follow highly varied pathways, such that there are few common meaningful patterns that characterize their course-taking.
– “Likely transferable” STEM courses appears to be the most important factor affecting students’ transfer outcomes in STEM. Students who earned less ‘‘likely transferable” STEM credits (say, 12-24 credits), but took quite a few math courses and a good amount of STEM credits would also have an increased probability of transfer in STEM.
– Provided strong exposure to transferable STEM courses, women seem to only need a minimal amount of math credits in order to succeed in the STEM transfer pathway, and additional math beyond six credits does not add much to boost their chances of transferring in STEM.
– After accounting for course-taking patterns, the only outstanding demographic characteristics that may influence and intersect with course-taking patterns are gender and age.

 

Scholarship Types: Journal Article Reporting Empirical ResearchKeywords: CollegeCommunity CollegeCourse-Taking PatternsGenderSTEMTransferRegions: NationalMethodologies: QuantitativeResearch Designs: Secondary Survey DataAnalysis Methods: Data Mining Sampling Frame:Community College Students
Sampling Types: Nationally RepresentativeAnalysis Units: StudentData Types: Quantitative-Longitudinal
Data Description:

Beginning Postsecondary Students Longitudinal Study (BPS:04/09) and the Postsecondary Education Transcript Study (PETS:09). This study follows a nationally representative, first-time postsecondary beginning cohort in 2003-2004, BPS:04/09 contains survey data at three points in time: in respondents’ first year of college, and then again three and six years after they started postsecondary education. Of critical importance to this study, transcripts were collected under PETS:09 from all 3030 eligible postsecondary institutions attended by the BPS respondents over a 6-year period.

The author restricted the study sample to beginning postsecondary students at community colleges who took and passed at least one non-remedial STEM course during the first year of postsecondary attendance. The author further limited the sample to students who may hold the intent to transfer. After applying these criteria, the final sample of 2330 students out of the nearly 5550 BPS panel respondents who began at a public 2-year institution.

The demographics controlled for are: gender, age, race, income (by quartiles), being a single parent, first-generation college student, English as 1st language, and high school GPA rank.

The DVs were: Transfer to a 4-year STEM major, transfer to a 4-year non-STEM major, and not transferred.

Key IV: course classifications include the following areas: (a) ‘‘likely transferable’’ STEM courses;
(b) ‘‘likely terminal’’ STEM courses; (c) mathematics courses within CIP 27, except for
those designated as remedial math; (d) English courses within CIP 23, except for those
designated as remedial English; (e) remedial courses; and (f) other.

Author utilizes ‘‘frequent pattern/association rule data mining’’ to interpret data. Utilizes A priori algorithm, decision list algorithm, and decision tree algorithm to determine course-taking patterns.

Theoretical Framework:
Relevance:Community College and STEM
Archives: K-16 STEM Abstracts