Browsing by Author "Logan, Kristi"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Exploring the Construct Validity of Principal Ratings as a Measure of Teacher Performance Using Meta-Analysis
    (2014-12) Logan, Kristi
    Multiple meta-analyses have been conducted exploring the relationships between subjective and objective performance, multi-source performance ratings, and ratings and personality. Rarely has this work included the occupation of teachers. This meta-analytic research explores three areas of teacher performance: (1) What is the relationship between principal ratings and student achievement scores? (2) What is the relationship between principal ratings and ratings completed by students, peers, parents, and other classroom observers, and the relationships between these ratings with one another? (3) What is the relationship between principal ratings and personality ratings completed either by the teacher (self-ratings) or others? Data was gathered from published and unpublished sources and analyzed using the Hunter and Schmidt (2004) psychometric meta-analytic method. The correlation between principal ratings and student gain was .17, similar to what has been found in the past (Medley and Coker, 1987). Value-added scores produced a greater relationship with principal ratings (r = .23) than when all gain results were used. Arithmetic tests (r = .24) exhibited the largest relationship with principal ratings. Multiple source rater data results found that principal ratings had moderate relationships between peers (r = .57), students (r = .31) and other classroom observers (r = .45). Student ratings exhibited their largest relationship with parent ratings, (r = .53). Self-ratings had low relationships with all other rating groups. Personality ratings generated either by teacher self-ratings or ratings by others using an overall prosocial personality factor correlated .28 and .45, respectively with principal ratings. Results using other ratings of personality showed that Conscientiousness and Emotional Stability are the most important for predicting high levels of performance (r = .23, for both). The results exploring the relationships between principal ratings and self-ratings of personality found low correlations for the global Big Five dimensions, ranging from -.11 to .06. Overall, this research provides new information about the relationships that principal ratings have with other criteria and predictors. It raises many questions that can be explored in future research.