angle-up angle-right angle-down angle-left close user menu open menu closed search globe bars phone store

Precollegiate Curricula, University Characteristics, and Field Persistence Among Science Majors

Pennock-Roman, Maria
Publication Year:
Report Number:
GREB-95-02R, RR-99-10
ETS Research Report
Document Type:
Page Count:
Subject/Key Words:
Academic Persistence College Science Computer Science Data Analysis Engineering Females Graduate Record Examinations (GRE) Graduate Record Examinations Board Graduate Study Institutional Characteristics Majors (Students) Males Mathematics Multiple Regression Analysis Scholastic Assessment Test Student Characteristics Undergraduate Study Whites


Few studies have attempted to examine simultaneously how student and university characteristics are associated with the persistence of women and men in the sciences. The present study took advantage of a unique data set containing the records of individual students on the SAT (1979-1982) and their corresponding records on the GRE (1985-1990), including the questionnaires that accompany each test. Information on the students' undergraduate institutions from the College Board Annual Survey of Colleges was then added to these records. The data set included 48,492 non-Latino White students with English as their best language who indicated an interest in computer, engineering, mathematical, or physical (CEMP) science fields in one or more responses to four questions about their intended major field of study. Two transition points were studied: from intended college major as high school seniors to actual undergraduate major (Stage I), and from actual undergraduate major to intended graduate school major (Stage II). Persistence was defined as staying within the broad category of CEMP sciences. The descriptive findings showed net losses from the CEMP pool of students at both transition points and greater field switching for women both into CEMP sciences and out of CEMP sciences. The predictive findings for both transitions, using logistic regression with random coefficients, showed that student and university characteristics (entered jointly) were associated with field persistence for the majority of fields. Significant gender differences in field persistence remained for most, but not all, fields after controlling for student and university.

Read More