The Development of a Model for Behavioral Observation Performance of Instructional Personnel
Feltner, Raymond Lawrence
MetadataShow full item record
The problem. The purpose of the project was to develop a performance evaluation model that would measure teaching behaviors of instructional personnel on a systematic and objective basis. The model was intended to reduce subjectivity and bias and yield high reliability and agreement among raters. Procedure. There were two urban and one rural school districts that participated in field testing the model for evaluating instructional personnel. After the instruments were developed, each district field tested the model by actual classroom observation of teaching activities. Twenty-two teachers from kindergarten through twelfth grade and special education were selected by school principals to be observed. Each observation for field testing and collecting data for the model was conducted by observers in teams of three. After the data were returned, they were analyzed for correlation of coefficient to determine reliability between raters. A percentage of agreement among raters was also determined from the data. Findings. The analysis provided inter-rater reliability of .90 for sixteen of the nineteen classroom observations. Mean inter-rater agreement of 85 percent was found on sixteen of the nineteen observations. Results of the three field sites were quite similar, Indications were the size of school district or grade taught did not influence the results. Conclusions. The results of the project provide support for the statement that a model can be developed for evaluating instructional personnel with high reliability and interrater agreement. The conclusions can be drawn: (1) Instructional personnel can be objectively and reliably observed; (2) Principals can be trained to observe and accurately identify teaching behaviors; (3) Previous studies of behavior observation methods have been supported by this project. Recommendations. Recommendations included: (1) School districts planning to replicate the model emphasize the procedures for training; (2) Instructional personnel be included in any design or modification of a performance evaluation model; (3) A five point, rather than a three point, scalogram be developed for each behavior scale; (4) Greater representation of the performance evaluation be obtained through reviews by educational administration preparation programs.
x, 137 leaves. Advisor: Richard H. Lampshire