Date of Award


Degree Type


University or Center

Atlanta University (AU)

Degree Name


Education and Administration

First Advisor

Dr. Ganga Persuad


This study examined whether or not the school's mean reading score could be explained by the teachers' perceptions of the principals' supervisory behaviors and teachers' characteristics and whether each student's reading score can be explained more by the principals' supervisory behaviors or by the students' perceptions of the fifth grade class climate or by the student's selected biographic variables in a large metropolitan school system.

Ten elementary schools were selected by experts in the instructional department of a large metropolitan school system. Nine of the selected schools were Project Achievement Schools in which the students scored below the national norms on the Iowa Tests of Basic Skills (ITBS). The students and teachers samples consisted of one hundred and seventy-eight regular classroom teachers and four hundred and twenty-five fifth grade students. The instruments used were the Student Perceptions Questionnaire extracted from the Teacher Performance Assessment Instrument developed by the State Department of Education and the Instructional Supervisory/Behavior Questionnaire developed by Dr. Ganga Persaud. The student achievement test used was the 1985 Iowa Tests of Basic Skills (Level 10) mean reading scores of individual students and mean school reading score for each of the selected schools.

In a regression analysis of the data, teachers’ degree qualifications, experiences, and perceptions of principals’ supervisory behaviors, in that order, predict the school mean reading score. The principals’ supervisory behaviors correlated inversely with the mean reading score of the school. The overall variance, however, is small - approximately 7 percent.

In a regression analysis of the data, students' perceptions of the class climate and the principals' supervisory behaviors, in that order, predict students' reading scores. The relationships are inverse for both variables with reading scores indicating consistency between the teachers and students' perceptions. The selected biographic variables make smaller but insignificant contributions to the students' reading scores. The overall variance, however, predicted for all variables is small just over 15 percent.

The results support the Edmonds' and Lezotte's Schools for low achievers. Support was not found for the Coleman studies that socio-economic and environmental variables were more important than the school variables for student achievement. This was probably due to the bias of the sample in favor of low socio-economic status (SES).