Abstract
It is known that when a previously-presented stimulus is presented again, its duration is estimated to be longer than in the initial presentation (Witherspoon & Allan, 1985). Such a prior presentation effect has been attributed to enhanced perceptual processing due to perceptual priming. We used the DRM procedure (Deese, 1959; Roediger & McDermott, 1995) to examine whether this effect can be mediated by strictly conceptual processes without any perceptual repetition. Specifically, we asked whether the duration of a word that has never been presented but that elicits false memory (recall/recognition) is judged to be longer than a control word that does not elicit false memory. If the prior presentation effect can be mediated by conceptual rather than perceptual processes, then the estimated duration of a falsely recognized word should be longer than that of a control word. The present experiment consisted of three phases. In the learning phase, subjects memorized several lists of sequentially presented words. The words in each list were all converging associates of a lure word that was not presented. In the temporal production phase, subjects saw a word and pressed a key as soon as they judged that the word had been on display for 2.5s. The words were a) old words that had been presented in the learning phase, b) lure words, or c) control words that had never been presented. In the recognition phase, subjects judged whether or not they had seen these three types of words in the learning phase. As expected, prior presentation increased the estimated duration: the temporal production of the words that had been presented was shorter than that of the control words (Temporal production bears an inverse relationship to the estimated duration). More importantly, the estimated duration of falsely recognized words was longer than that of a control word. The results suggest that the prior presentation effect on temporal estimation can be mediated by conceptual process without any perceptual repetition.