Release from output interference in recognition memory: A test of the attention hypothesis

Abstract

Retrieval both helps and harms episodic memory. For example, long-term memory can benefit more from testing memory than from additional study (e.g., Roediger & Karpicke, 2006a, 2006b) However, the benefits of testing are far from ubiquitous, and memory is even harmed by testing in some cases (Malmberg, Lehman, Annis, Criss, & Shiffrin, 2014). For instance, accuracy decreases across a series of episodic memory test trials, known as output interference (OI, for example,

The effect of perceptual information on output interference

“…Interference from these features builds up as they are encoded during test. This is similar to our explanation for release from OI observed for faces (when studied with words; Criss et al, 2018). What is perhaps surprising is that the rebound appears to be complete, or, performance returns to the same level as the first test bin.…”

“…This would cause any test cue to be remembered less well as trials (or time) pass, reducing performance. While reasonable, this account fails to explain release from OI (Criss et al, 2018;Malmberg et al, 2012). On the other hand, context drift may be due to external factors related to the stimulus, such as perceptual form.…”

“…OI was observed in eight out of 12 conditions. In no case was a release from OI observed when perceptual form of (Criss et al, 2018;Malmberg et al, 2012) and suggests that perceptual information does not contribute to OI. However, OI was not observed in several conditions, making the results difficult to interpret.…”

“…OI is present if slope is negative. To evaluate whether there is a release from OI when the perceptual form of test items changed, we conducted a 2 (bin4/ bin5) × 2 (mixed/blocked) ANOVA (following Criss et al, 2018;Malmberg et al, 2012). Release from OI is present if OI was detected and accuracy in Bin 5 is higher than accuracy in Bin 4 in the blocked condition.…”

“…A second potential account of OI is the attention hypothesis (e.g., Criss, et al, 2018;Watkins & Watkins, 1975;Wickens, 1970), according to which attention decreases with increasing test trials, and the observed decrease in accuracy is due to lack of effort, attention, or motivation. A release from OI may be observed when changing stimulus classes, conceptually or perceptually, boosts attention and, therefore, accuracy.…”