An analysis of mentor-based induction program components and teachers' perceptions of improvement in teaching

This study examined the relationship between mentor-based induction program components and beginning teachers' self-perceived improvement of their teaching as a result of work with their mentors. The conceptual framework for this study was adapted from a National Center for Education Evaluation report by identifying key components of mentor-based induction programs specified in the literature and contextual factors that are useful to consider when interpreting analyses. A restricted data license for the First through Third Waves Beginning Teacher Longitudinal Study Preliminary Data File was obtained from the National Center for Educational Statistics for the purpose of this study. Given that the outcome variable is ordinal, hierarchical generalized linear modeling (HGLM), the non-linear counterpart to hierarchical linear modeling was used. HGLM results indicated there is not a statistically significant relationship between beginning teachers' self-perceived overall improvement of their teaching as a result of working with their mentors and mentor criteria, mentor collaboration time, observation time, and the frequency with which the mentor teacher worked with the beginning teacher in eight identified areas.