The purpose of this study was to determine if a relationship existed between exam grades and students' causal attributions of their grades in a developmental mathematics course at a community college. Also investigated were differences in causal attributions of grades between Traditional and Nontraditional students. In addition, among Nontraditional students, differences based on gender were examined. The sample consisted of 331 completed questionnaires from 24 sections at a southern community college in the Spring 2010 semester.

The instrument used was a self-report questionnaire consisting of four parts: (a) demographic data section; (b) seven questions to determine students' classification; (c) short answer section about students' exam grade, and an attribution for the exam grade; (d) Revised Causal Dimension Scale (CDSII). A Pearson chi-square test was conducted between low-graded and high-graded students to test for a relationship between exam grade and reported attributions. Multivariate Analysis of Variance (MANOVA) was performed between student classifications, based on exam grade and scores on the CDSII to test for any relationships. Multivariate Analysis of Variance was also performed between gender of Nontraditional students and scores on the CDSII.

The statistical analysis indicated a difference in reported attributions of low-graded and high-graded students. Low-graded students' reported attributes were spread across the eight categories while high-graded students attributed Internal-Stable-Controllable and Internal-Unstable-Controllable attributes most frequently. This overall trend appeared in all student classifications but Minimally Nontraditional students. Reported attributes for this group were scattered over the eight categories regardless of exam grade.

On the CDSII, neither low-graded students nor high-graded students showed significant differences in Locus of Causality or Stability dimensions when distinguished by student classification. For low-graded students, there was a significant difference in the Personal Controllability dimension. For high-graded students, a significant difference appeared in the Personal Controllability dimension and the External Controllability dimension. When compared by gender, low-graded Nontraditional students differed on the CDSII in the Locus of Causality dimension, with females attributing their grade more towards internal traits as compared to males. Among high-graded Nontraditional students, there was no significant difference in any of the dimensions.