The Role of Semantic Relatedness in the Pretesting Effect

The pretesting effect is the finding that attempting to make a guess at the target before it is revealed is more beneficial than only studying the cue-target pair. This effect is typically obtained in semantically related word pairs, but not in unrelated pairs. Experiment 1 examined the pretesting effect when the unrelated cue and target were embedded in a sentence. In the sentence condition, subjects were presented with a sentence frame and attempted to guess a word that fit into the sentence, then studied the cue-target pair with the sentence. In the word-only condition, subjects were asked to guess a word related to the cue, then studied the cue-target pairs alone. There was a pretesting effect when the unrelated cue-target pair was embedded in a sentence, but not when they were presented alone. Experiment 2 showed that there was also a pretesting effect when the cue was presented without a sentence frame during pretest and the cue-target pair was presented with a sentence during study. Experiment 3 showed that the pretesting effect was eliminated when subjects were explicitly instructed to connect the cue-target pair on the subsequent study trial. Overall, the results challenge accounts of the pretesting effect that are based on semantic relatedness or activation of the target.