Assessing the Skill of Variable Identification with Multiple Choice Items

Increased emphasis on public school accountability has resulted in several states including a science component in their elementary and middle school testing program. This component often includes items designed to measure process skill ability. The purpose of this study was to examine the performance of high school students on four types of multiple choice items used to assess student's ability to identify manipulated and responding variables. Items differed only in the stimulus material presented to students. Stimulus material was either a question focusing on the relationship between two variables, a hypothesis, a description of an experiment, or a description of results of an experiment. Student performance on these item types was compared to their performance on a standard Piagetian interview task of variable identification. The results of the study included: (1) the "question" format was the most difficult, while the "results" format appeared to be the easiest; (2) the "description" item type had a higher correlation with the total interview than any other item type. Among conclusions reached in this study was that although all four item types operate similarly, they did not correlate very highly with the interview/observation task. (CW)