Can Practice Calibrating by Test Topic Improve Public School Students\u27 Calibration Accuracy and Performance on Tests?

The effect of a calibration strategy requiring students to predict their scores for each topic on a high stakes test was investigated. The utility of self-efficacy towards predicting achievement and calibration accuracy was also explored. One hundred and ten sixth grade math students enrolled in an urban middle school participated. Students were assigned to either a calibration practice group or a no practice condition. Students in the practice condition completed a self-efficacy scale specific to math at the beginning of the study. They also practiced making predictions for each topic on each of three tests over a three month period to determine if their calibration accuracy and performance on tests would be increased. Students in both the practice and no-practice conditions calibrated their scores topically on the final, high stakes math test at the end of the course. There was not a significant difference between the conditions in calibration accuracy on the final, high stakes test, indicating that calibration practice did not improve accuracy. There was no significant difference between the practice and no practice conditions in on achievement. However, a significant relationship was found between achievement level and calibration accuracy. Higher achieving students in both the calibration practice and no practice conditions were significantly more accurate then lower achieving students in both conditions. Self-efficacy was not found to be predictive of achievement or calibration accuracy. Further research is needed to identify more effective strategies for enhancing metacognitive judgments, self-efficacy, and performance