Sleep can reduce proactive interference

Abstract

Sleep has repeatedly been connected to processes of memory consolidation. While extensive research indeed documents beneficial effects of sleep on memory, little is yet known about the role of sleep for interference effects in episodic memory. Although two prior studies reported sleep to reduce retroactive interference, no sleep effect has previously been found for proactive interference. Here we applied a study format differing from that employed by the prior studies to induce a high degree of proactive interference, and asked participants to encode a single list or two interfering lists of paired associates via pure study cycles. Testing occurred after 12 hours of diurnal wakefulness or nocturnal sleep. Consistent with the prior work, we found sleep in comparison to wake did not affect memory for the single list, but reduced retroactive interference. In addition we found sleep reduced proactive interference, and reduced retroactive and proactive interference to the same extent. The finding is consistent with the view that arising benefits of sleep are caused by the reactivation of memory contents during sleep, which has been suggested to strengthen and stabilise memories. Such stabilisation may make memories less susceptible to competition from interfering memories at test and thus reduce interference effects.

Keywords:

• Proactive interference
• Retroactive interference
• Sleep
• Delay
• Forgetting

The project was conducted as part of Magdalena Abel's doctoral dissertation.

The project was conducted as part of Magdalena Abel's doctoral dissertation.

Notes

¹ In literature on interference effects a distinction is frequently made between single-list conditions controlling for either retroactive or proactive interference (e.g., Anderson & Neely, 1996). This distinction pertains to the exact point of time at which single-list encoding takes place. It is reasoned that retroactive interference is adequately assessed when single-list encoding takes place at the same point of time as encoding of the first list in interference conditions, whereas proactive interference is adequately assessed when single-list encoding takes place at the same point of time as encoding of the second list in interference conditions. We also differentiated between the two single-list encoding conditions, and half of the participants in all no-interference conditions encoded the list as control for retroactive interference, while the other half encoded it as control for proactive interference. However, because comparisons between the two single-list encoding conditions never revealed significant differences in recall level in any of the delay conditions, all ts(47)<1.30, all ps>.200, we pooled the data for the two single-list encoding conditions and report the results without further mentioning the distinction.

² Retroactive and proactive interference after the 12-h delays were analysed by means of two separate 2×2 ANOVAs, because interference was manipulated between participants in these conditions. As participants studied either the single list or both interfering lists, no all-encompassing 2×3 ANOVA could be calculated.

³ A 2×2 ANOVA comparing the 12-h delay conditions only showed that there was no significant difference between conditions, and no significant interaction, all Fs(1, 46)<1.0.