Assessment of a mentor program on self-concept and achievement variables of middle school underachievers

The increasing focus on the underachiever has intensified the search for affective education models. Underachievement is frequently associated with a low self-concept. Current studies are sparse, indicating that Mentor Programs may improve self-concept, but empirical assessments are lacking.

This study investigated the efficacy of a mentor model on self-concept and achievement variables of intermediate school underachievers.

A Mentor Program model was implemented with an experimental group of 55 underachieving students in a Fairfax County, Virginia, intermediate school. A 42 student control group of underachievers in another Fairfax County intermediate school were monitored. Forty education staff members served as mentors to the experimental group of students. The study was of a quasi-experimental, nonequivalent control group design.

Primary measurement instruments used were the Self-Concept and Motivation Inventory (SCAMIN), an appropriate measure of self-concept in the school setting, the Grade Point Average (GPA), the standard measure of academic achievement, and the Failure Rate, including students retentions and classes failed.

Four research questions were investigated. For testing overall effects of the treatment/Mentor Program at the school level, a Value Added Analysis was performed. For testing the hypotheses, the following analyses were undertaken: ANCOVAs were performed on the achievement data; t-tests and ANOVAs were performed on the self—concept data, Chi-square, t-test, and ANOVA were performed on the failure data. Canonical Correlation Analysis was performed to explain the relationship between the predictor measures and the criterion measures. Descriptive and ethnographic information in the form of quantitative and qualitative data analyses added to the breadth of the assessment.

Results revealed that the Mentor Program produced positive, nonsignificant gains at the experimental school. The gains were better than those at the control school, but not significantly better. Analysis of the results also disclosed changes in the study design should be considered for future research. Recommendations include two year assessments, multiple school comparisons, and longitudinal studies.

Post program results from teacher ratings, mentor and students evaluations were positive, providing qualitative statements of program worth. The findings and conclusions drawn from this study serve to further improve program evaluation and assessment of Mentor Programs.