The purpose of this study was to compare the effects of three testing procedures on comprehensive final examination performance. The 64 undergraduates used as subjects were randomly assigned to one of three conditions: a non-repeated single-unit, a non-repeated cumulative, or a repeated single-unit testing procedures. Subjects in the non-repeated testing procedures were administered a weekly multiple-choice test taken from content covering either (a) the most recent material or (b) all the material to-date. Subjects in the repeated testing procedure were administered up to three alternate forms of a multiple-choice test taken from content covering the most recent material. The dependent measure was a comprehensive 100 item multiple-choice test. Grade point average was used as a covariate.An Analysis of Covariance was applied to determine whether there was a significant difference between (a) the means of the non-repeated cumulative and the repeated single-unit procedures, and (b) the average mean of these two procedures and the mean of the non-repeated single-unit procedure. The .01 level was used for judgments about statistical significance.Findings1. No significant difference was found between the mean final examination score obtained by the students in the non-repeated cumulative and the repeated single-unit testing groups after the effects of the covariate were partialled out.2. No significant difference was found when the average of the final examination scores obtained by the non-repeated cumulative and the repeated single-unit groups was compared to the mean final examination score obtained by the non-repeated single-unit group.3. The subjects in the repeated single-unit testing group obtained a significantly greater number of points on the examinations administered throughout the quarter.Conclusions1. Previously reported differences between repeated versus non-repeated and repeated versus cumulative testing on final examination performance may be reflecting only differences in testing frequency.2. The higher grades obtained by students given multiple opportunities to "master" material may not reflect actual end-of-course achievement by a comprehensive final.3. To aid in the valid interpretation, generalizability, replication and synthesis of research findings, detailed descriptions of the independent and dependent variables used in testing procedures research seem warranted.
| Identifer | oai:union.ndltd.org:BSU/oai:cardinalscholar.bsu.edu:handle/180213 |
| Date | January 1984 |
| Creators | Rohm, Rosalie Ann |
| Contributors | Miller, Ebert L. |
| Source Sets | Ball State University |
| Detected Language | English |
| Format | 100 leaves : ill. ; 28 cm. |
| Source | Virtual Press |
Page generated in 0.0021 seconds