Is it relevant? Influence of trial manipulations of prospective memory context on task interference

Abstract

Prospective memory (PM) research has often investigated if having an intention interferes with ongoing activities, but rarely by linking the intention to a particular context. We examined effects of trial-by-trial changes in whether the context (defined by colour) was relevant for the nonfocal PM task. The ongoing task involved speeded decisions about the position (left/right) of the upper-case letter in a pair, and the PM task consisted of pressing an additional key if the upper-case and lower-case letters were in a specified colour and the same letter. Trials switched between two colours either randomly or predictably in eight-trial blocks. We also manipulated the presence/absence of occasional same-letter pairs in the irrelevant context. Results showed higher cost of having a nonfocal PM task when ongoing stimuli matched than when they mismatched the target's colour. Moreover, cost for intention-irrelevant stimuli was minimized, though never eliminated, by blocking match/mismatch trials. These findings highlight the role that local changes in intention-related context play in task interference and support a view of monitoring as a flexible mechanism. Additionally, the study introduced a novel way of embedding intention-related events in the irrelevant context shortly before the occurrence of PM targets, with results tentatively suggesting that such events might impair target detection.

Keywords:

• Prospective memory
• Context
• Task interference
• Lures
• Preparatory attention

Notes

¹ The colour (match, mismatch) factor was omitted here as the distinction between match and mismatch trials is linked directly to the PM task (i.e., colour associated with the PM targets determines the trials that fall into each colour category) and is therefore not applicable to the control condition.

² Note that this measure of cost underestimates actual task interference because it does not account for practice effects on the ongoing task from the baseline to the PM phase. Thus, slowing in the PM conditions contrasted with an 18-ms practice effect in the no-PM-demands control condition.

³ This finding is noteworthy because task interference has been previously identified as supporting PM retrieval with nonfocal targets (e.g., Scullin, McDaniel, & Einstein, 2010). Thus, we further investigated whether PM accuracy was related to task interference for intention-relevant trials. We conducted a 2 × 2 analysis of covariance that included condition (random, blocked) and lures (present, absent) as between-subjects factors and task interference in colour match trials as a covariate. There was a marginally significant effect of the covariate, F(1, 77) = 3.45, MSE = 0.07, p = .067, η²_p = .04, and the effect of lures was now only marginally significant, F(1, 77) = 2.98, MSE = 0.07, p = .088, η²_p = .04. Thus, the negative effect of out-of-context lures on target detection was reduced after controlling for differences in ongoing task slowing in colour match trials.

⁴ For completeness, we conducted a similar 2 × 2 × 8 mixed ANOVA for RTs in colour match trials (see left-hand panel of Figure 3). There was a significant main effect of trial, F(7, 266) = 3.21, MSE = 1288.94, p = .003, η²_p = .08, as RTs were slower to the first than to the following trials in the sequence. The phase by trial interaction was only marginal, F(7, 266) = 1.83, MSE = 1281.30, p = .082, η²_p = .05, and reflected a trend for slowing at the beginning of the sequence of trials to be more evident in the PM than baseline phase.

⁵ In the blocked condition, lures/lure controls were always presented as the fifth item in the sequence of eight colour mismatch trials. Therefore, in this condition the depicted trials all represent colour mismatch trials with the exception of lure + 4, which was the first item of the sequence of eight colour match trials that followed.