Testing Imagery and Free Recall: Alternate Forms for Serial Assessment

Abstract

Three alternate forms of a 100-item free-recall memory test were developed. Each form contained 50 readily visualized words and 50 words difficult to visualize. The performance of subjects retested weekly on the same test form was compared with the performance of subjects retested weekly using alternate forms. 32 undergraduate students were tested once with each form in a partially counterbalanced design. Another 12 students were tested three times using the same test form. The three test forms showed significant parallel-form reliability. The performance of people retested with the same form improved significantly more across sessions than was noted in groups given alternate forms. The importance of parallel forms in serial testing and implications for clinical assessment are discussed.