A pilot study testing effects of feedback on teaching behavior as measured by student ratings at New River Community College, Dublin, Virginia

Research Problem. The focus of this study was utilization of feedback which enabled teachers to modify teaching behavior after receiving the feedback.

Research Methodology. The population was seventysix full time and part time instructors employed at New River Community College, Dublin, Virginia, during the Spring quarter of the academic year, 1974-75. Six instructors were assigned to each of two control groups, three treatment groups, and a posttest-only group, using an equal-size stratified random sampling technique. The research design was a modified Solomon Four-Group.

Pretests and posttests (Purdue Rating Scale for Instruction) were administered to all groups except the posttest-only group.

Test of the Hypotheses. Data collected from the pretests and posttests were used to test two hypotheses. Hypothesis one, no significant difference in teaching behavior measured by student ratings shall occur between student ratings administered before and those administered after feedback is given the teacher about his teaching performance, was tested at the .05 alpha level using a multivariate T² test. The hypothesis was not rejected.

Hypothesis two, no significant difference in teaching behavior measured by student ratings shall occur among teachers given different types of feedback, was tested at the .05 alpha level using a multivariate analysis of variance and linear contrasts. The hypothesis was not rejected.

Results.

1. No significant change occurred in teaching behavior as measured by student ratings after feedback was administered.

2. No specific type of feedback was associated with behavioral changes.

3. No significant change occurred in teaching behavior during the ten-week quarter.

4. There appeared to be no relationship between age of instructors, educational level of instructors and occupational experience other than teaching and any one of the instrument items of the Purdue Rating Scale.

5. There appeared to be a relationship between teaching area and three instrument items.

6. There appeared to be a relationship between teaching experience and three instrument items.

Conclusions.

1. Pretest scores were higher than anticipated apparently making subsequent posttest scores not significantly different.

2. Student familiarity achieved after a two-quarter exposure to teachers might have resulted in students overlooking undesirable teaching behavior.

3. Failure to correlate students' and teachers' perceptions of teaching behavior may have eliminated commonalities needed for teachers to make self-assessments using student ratings.

4. Lack of association of specific feedback with behavioral changes appeared to result from failure of teachers to utilize the combination feedback appropriately.

S. One ten-week quarter appeared insufficient time for teachers to change behavior needing change.

Recommendations. Based on conclusions it was recommended that research be conducted investigating:

1. Possibilities that demographic characteristics of instructors are related to results of student ratings.

2. The likelihood that student ratings are skewed positively making subsequent positive changes in teaching behavior difficult to measure.

3. The use of more than one academic quarter to conduct the research.

4. Effects of teacher demographic characteristics on teaching behavior.

5. Comparison of effects of feedback and demographic characteristics of teachers on student ratings.

6. The correlation between teachers' and students' perceptions of positive teaching behavior.