Is metamemory implicit? evidence of priming and antipriming in individuals with and without traumatic brain injury.

Abstract

Implicit memory refers to the phenomenon of prior exposure to a stimulus influencing cognitive processing on subsequent exposure to that stimulus, irrespective of one's awareness of such influence. Metamemory refers to the presumably explicit act of making judgments about one's memory. Theories vary regarding possible relationships between implicit memory and metamemory, with mixed findings of previous research. Some researchers argue that there is no relationship, that metamemory does not rely on input from implicit memory only on explicit memory (Hart, 1965; Koriat, 1995), while others argue that metamemory is driven only by implicit processes (Reder & Schunn, 1996). Still others argue that the metamemory system relies on input from both implicit and explicit memory (Vernon & Usher, 2003). To date, no study has examined whether subliminal masked priming and antipriming can influence individuals' metamemory judgments of learning (JOLs). The present study independently varies subliminal masked priming conditions (using baseline, prime, and antiprime conditions) and observes the influence on participants' immediate and delayed JOLs and relative predictive accuracy, in a paired associate learning task. This study also examines long term repetition priming and antipriming (measured in a visual object identification task) to determine whether the magnitudes of these effects correlate with those found in the paired associate learning task. Both neurologically normal individuals and those with traumatic brain injury (TBI) participated in this study. Metamemory deficits are common after TBI. Thus, inclusion of this clinical population in the present study is critical in evaluating whether or not such individuals' metamemory judgments can be influenced by, or correlated with, priming or antipriming. Furthermore, to date, no study has examined antipriming among TBI survivors; therefore their participation in the visual object identification task will shed light on antipriming in that population, and will provide a perceptual implicit memory benchmark against which to compare any findings of implicit metamemory. Results demonstrate no main effect of subliminal masked priming and antipriming on participants' metamemory judgments for JOL ratings, predictive accuracy (gamma correlation), or response times for judgments of the likelihood of future recall of target words. However, there was a trend towards an interaction effect of priming and JOL timing (i.e.: immediate vs. delayed), such that JOLs made immediately after study were higher for the primed items than for the baseline or antiprimed items. In contrast, antipriming did significantly influence explicit recall, with slowing of explicit recall response times for antiprimed target words, relative to word-pairs in the baseline condition. Difference scores between JOL ratings and recall accuracy reflected overconfidence among TBI survivors, and good calibration among controls. There was a significant decrease in the overconfidence of TBI survivors for antiprimed items. However, much of this lowering was caused by an increase in recall accuracy for antiprime items. In the visual object identification task, antipriming was robustly observed for both participant groups; there was significant slowing observed for responses to antiprimed items. Finally, a mild positive correlation between the degree of priming for visual object identification and the degree of priming of relative predictive accuracy in the paired associate learning task was found across participants. The present research finds weak evidence for the role of implicit memory, in the form of masked priming and antipriming, in JOLs and relative predictive accuracy in paired associate learning. Future research providing masked priming stimuli immediately prior to the solicitation of JOLs may address possible attenuation of such effects which may have occurred in the present study. Additionally, using numbers from the JOL rating scale as subliminal stimuli may shed additional light on the type of representation taken as input by the metamemory system. Finally, exploration of long term priming and antipriming of metamemory in the TBI population should be undertaken to determine whether or not there is a priming benefit to metamemory judgments through supraliminal implicit memory.