Perceptual degradation of visual stimuli decreases performance in many tasks. In word-reading, response time (RT) for words with no blur (NB) are slightly faster than for words with low blur (LB) and much faster than for words with high blur (HB). However, subsequently probing recognition memory for these words reveals superior memory sensitivity for HB words than NB words, and numerically worse memory for LB words than NB words. (Rosner, Davis & Milliken, 2015). This result suggests that a high level of perceptual degradation can enhance long-term memory, perhaps due to the upregulation of attention in response to processing difficulty at the time of encoding. Borrowing from the literature on pupil dilation as an index of mental effort (e.g., Beatty, 1982), we incorporated pupil size as a measure of attentional engagement in the present study. In the encoding phase, half of the participants were presented with NB and LB words intermixed, and the other half with NB and HB words intermixed. In the test phase, participants completed a surprise recognition memory task. Pupil size was recorded throughout the experiment. As expected, RT in the encoding phase increased with stimulus blur. More important, recognition memory in the test phase (relative to NB words) was slightly worse for LB words and significantly better for HB words. Evoked pupillary response (EPR) during the encoding phase did not differ between NB and LB words, but was larger for HB than NB words. Critically, the larger EPR to HB words at study was driven by 'old' words that were later recognized (hits), rather than those that were not (misses). Follow-up analyses showed the EPR results did not depend on longer time-on-task (slower RT) for HB words. The results are consistent with an attentional-upregulation account of the effect of perceptual degradation on long-term memory.

Meeting abstract presented at VSS 2018