Modeling criterion shifts and target checking in prospective memory monitoring.

Abstract

Event-based prospective memory (PM) involves remembering to perform intended actions after a delay. An important theoretical issue is whether and how people monitor the environment to execute an intended action when a target event occurs. Performing a PM task often increases the latencies in ongoing tasks. However, little is known about the reasons for this cost effect. This study uses diffusion model analysis to decompose monitoring processes in the PM paradigm. Across 4 experiments, performing a PM task increased latencies in an ongoing lexical decision task. A large portion of this effect was explained by consistent increases in boundary separation; additional increases in nondecision time emerged in a nonfocal PM task and explained variance in PM performance (Experiment 1), likely reflecting a target-checking strategy before and after the ongoing decision (Experiment 2). However, we found that possible target-checking strategies may depend on task characteristics. That is, instructional emphasis on the importance of ongoing decisions (Experiment 3) or the use of focal targets (Experiment 4) eliminated the contribution of nondecision time to the cost of PM, but left participants in a mode of increased cautiousness. The modeling thus sheds new light on the cost effect seen in many PM studies and suggests that people approach ongoing activities more cautiously when they need to remember an intended action.

The importance of age-related differences in prospective memory: Evidence from diffusion model analyses

“…Horn and Bayen (2015) similarly found that PM demands produced changes in threshold, but not drift rate, but additionally found that variables thought to influence the degree of monitoring enacted (i.e., importance of intention, cue frequency, and cue focality) selectively increased nondecision time. Together these results suggest that PM costs may reflect both response caution and target-checking, as indicated by changes in boundary separation and nondecision time, respectively, rather than decreased processing efficiency (i.e., drift rates).…”

“…Participants performed an ongoing lexical decision task with a nonfocal intention (monitor for the “tor” syllable) and were either instructed that it was more important to detect all the PM cues (PMI condition), or that it was more important to do well on the ongoing task (ongoing task importance, or OTI, condition). Prior research with young adults has demonstrated that PMI instructions generally increase cue detection at the cost of ongoing task performance (Horn & Bayen, 2015; Loft et al, 2008). However, the influence of importance on PM in the context of aging is mixed, with one study showing a selective benefit in cue detection from PMI instructions for older adults (Hering et al, 2013), one showing a benefit for both age groups (Smith & Hunt, 2014), and one showing no effect for either age group (Kliegel, Martin & Moor et al, 2003).…”

“…Hering et al suggested that the selective cue detection benefit for older adults in the PMI condition reflects that older adults may have “withdrawn part of the executive processing resources from the ongoing task to allocate them instead towards the fulfillment of the intended action” (p. 77), which would predict changes in the drift rate parameter had the authors implemented diffusion model analyses. However, Horn et al (2013) found no evidence for processing efficiency differences between young and older adults during a standard nonfocal PM task (i.e., no importance manipulation), and Horn and Bayen (2015) found that PMI instructions selectively influenced nondecision time (with younger adults). Thus, it is possible that other processes (e.g., target-checking, response caution) may have contributed to the findings of Hering et al, and may underlie age-related cue detection differences more generally.…”

“…Alternatively, delay theory posits that PM demands should produce changes in boundary separation because participants strategically increase response thresholds to allow more time for PM response selection to occur (Heathcote, Loft, & Remington, 2015). Lastly, Horn and Bayen (2015) suggested that PM demands may produce increases in nondecision time due to strategic checking of stimuli for intention-relevant details (i.e., target - checking ) either prior to or following the ongoing task-decision process. It is important to note that although target-checking is attentionally demanding (Guynn, 2003), in the context of diffusion modeling it is considered to be a serial process occurring before/after the ongoing task decision and therefore should not slow response selection by diverting attention away from the ongoing task.…”

Prospective Memory in Older Adults: Where We Are Now and What Is Next

“…For example, by investigating parameters from the ex-Gaussian distribution in younger adults, Loft et al [32] revealed a shift in the entire response time distribution (μ) in a nonfocal compared to a focal PM condition (indicating a more continuous PM monitoring profile in nonfocal PM) and an increase in skew (τ) in the nonfocal compared to the focal condition (indicating lapses of attention in nonfocal PM). Horn and Bayen [33] showed with the diffusion model approach that increases in peripheral nondecision time emerged in a nonfocal (but not in a focal) PM task, possibly reflecting a PM-target-checking strategy before and after the ongoing decisions. In addition, comparing IIV in younger and older adults, Horn et al [34] found that such increases in peripheral nondecision time in a nonfocal PM task were larger for older than for younger adults, possibly indicating older adults' lower capacity to recruit additional executive resources for PM-target checking).…”