The effect of immediate or delayed feedback, immediate retention test given after immediate and delayed feedback or omitted : amount of time spent with feedback, and verbal ability on long- and short-term retention of various levels of multiple-choice items and error pattern analyses

Abstract

This study investigated the most efficacious means of providing feedback to students on multiple-choice test items. Subjects were 135 undergraduate students enrolled at a mid-western university. Immediate feedback, defined as knowledge of correct results presented after the completion of the entire test, and delayed feedback, defined as knowledge of correct results presented 24 hours after the completion of the test, were examined. The presence or absence of an immediate retention test was also examined. The performance on the initial test, the students' verbal ability, and the amount of time spent with feedback were used as covariates. Also investigated was the relationship of the level of the test item (recall or complex) to the students' ability to profit from immediate or delayed feedback. The interference perseveration hypothesis was examined as a possible explanation of the delayed retention effect (DRE).Two analyses of the data were performed to test the DRE. The first was a univariate analysis of covariance of the four treatment conditions, utilizing the students' performance on the initial test, verbal ability and time spent with feedback as covariates. The results of this analysis were not significant. The second was a univariate analysis of covariance of the four treatment groups and the control group with the initial test score and the students' verbal ability covaried.The analysis yielded a significant treatment effect when the control group was compared to the four treatment conditions.The reliability correlation coefficients which were computed for the recall and complex test items were not adequate to be used in the analysis. Therefore the interaction between the level of the test items and the feedback condition was not tested.All of the students were asked to complete a pre- and postquestionnaire to assess their feedback preference and to determine if this preference changed over time. Results indicated the majority of students preferred to receive immediate feedback and little change was observed in their feedback preference.An item analysis was conducted in order to investigate the interference-perseveration hypothesis. The data did not support this theory.Conclusions were drawn and implications for further research were discussed.