How does semantic elaboration affect retrieval from episodic memory?

The semantic elaboration effect refers to the finding that semantic processing of words leads to better memory performance when the words are semantically congruent with the orienting questions with which they are studied. In Experiment 1, we were unable to replicate the congruency effect that motivated the concept of elaboration. We ran 8 follow-up experiments (Exp. S1–S8) that partially resolved our failed replication attempt: We found robust congruency effects in recognition tests, but not in free recall. This result can be explained by the hypothesis that elaboration improves memory by establishing effective retrieval cues, and elaboration is effective only inasmuch as these retrieval cues are available at test. To investigate the retrieval-cue hypothesis further, we asked whether participants are able to freely recall the orienting questions as retrieval cues. Recall of these questions was poor, providing a potential explanation for absent congruency effects in free recall. We formalized the retrieval-cue hypothesis in a simple computational model, applied it to Experiments 1 and S1–S8, and found that the model can account for the congruency effects we observed. In Experiment 3, we directly tested the models’ prediction that the size of the congruency effect is determined by the probability of recalling the orienting questions as cues by letting one group of participants learn the orienting questions before the experiment. We found a larger congruency effect in this group compared to a control group that did not learn the orienting questions, further supporting the retrieval-cue hypothesis of elaboration.