Literaturnachweis - Detailanzeige
Autor/in | Arnold, Carolyn L. |
---|---|
Titel | Using HLM To Investigate Instructor Grade Variability and Differences by Gender and Race-Ethnicity in Ethnically-Diverse Community College Math Courses. |
Quelle | (1996), (23 Seiten)
PDF als Volltext |
Sprache | englisch |
Dokumenttyp | gedruckt; online; Monographie |
Schlagwörter | College Mathematics; Community Colleges; Correlation; Females; Grades (Scholastic); Minority Groups; Multivariate Analysis; Student Characteristics; Teacher Characteristics; Two Year College Students; Two Year Colleges |
Abstract | A study was undertaken at a large, ethnically diverse community college in California to identify instructor and course factors associated with grade variability and grade differences by gender and race/ethnicity in mathematics courses. The study used a statistical procedure called hierarchical linear models (HLM) to analyze the relationship between within-course variables (i.e., average grades awarded and average grade differences by gender and race/ethnicity) and between-course variables (i.e., instructor and class characteristics). The study sample included all 2,440 students in 68 pre-college and college-level mathematics courses in one term, while student-level data for the HLM analysis included gender, race/ethnicity, and final grades. Instructor- and course-level data were instructor gender, race/ethnicity, experience, and part-/full-time status and the level of the mathematics course. Study findings included the following: (1) a significant difference in average grades that was found between classes was determined to be related to instructor experience, with instructors having 10 or more years of experience assigning an average of .5 grade points lower than instructors with less than 10 years experience; (2) no differences were found for grades received by gender within classes, although there was substantial variation between classes; and (3) minority students averaged .2 grade points below white students within classes, while no significant variance was found between classes. While the HLM method can allow new questions to be asked about factors affecting grades, the focus on instructor characteristics may raise sensitive issues when used within individual colleges. Contains 30 references. (HAA) |
Erfasst von | ERIC (Education Resources Information Center), Washington, DC |