Is it Harder to be Conservative in Recognition Memory? Exploring Shifts in Response Bias

Recognition memory can be described as identifying a stimulus as having been encountered before. There are individual differences in how judgments are made when encountering a stimulus with uncertainty as to whether or not it has been previously encountered. Response bias is the tenancy to choose one response over another, either erring on the side or saying the stimulus was encountered before (a "liberal" recognition bias) or erring on the side of saying the stimulus is new (a "conservative" bias). A conservative bias is often assumed to be more demanding than a liberal bias because maintaining a conservative bias requires more evidence when making a decision. The purpose of this study was to investigate whether it is more cognitively effortful to possess a conservative bias in recognition memory. Experiment 1 sought to examine whether divided attention would interfere with a participant's ability to engage in cognitive control, making response bias more liberal. If cognitive control is associated with a conservative bias, then dividing attention should cause participants to become more liberal. Second, I hypothesized that response reaction times would be slower for participants with a conservative bias, if they are utilizing more cognitive resources to make judgments with such a bias. Experiment 2 induced a liberal or conservative bias under either divided and full attention, with the prediction that accuracy would suffer under divided attention for participants with a conservative bias. Maintaining a conservative bias while dividing attention will require more cognitive resources, thus decreasing accuracy relative to when participants are induced to be liberal. The results did not support the idea that it is harder to possess a conservative bias. In Experiment 1, introducing divided attention did not cause participants to become more liberal in their response bias and reaction times were similar across both liberal and conservative biases. In Experiment 2, there was no difference in accuracy across participants induced to have a liberal or conservative bias. In addition, there was very little variation in reaction time between participants with a liberal, neutral, and conservative response bias. The results of this study suggest neither possessing a conservative or liberal bias require more cognitive control than the other. Rather, both conservative and liberal biases appear to be equal in the amount of cognitive control it takes to make a recognition decision.