Individual differences in working memory capacity predict benefits to memory from intention offloading

ABSTRACT

Research suggests that individuals with lower working memory have difficulty remembering to fulfil delayed intentions. The current study examined whether the ability to offload intentions onto the environment mitigated these deficits. Participants (N = 268) completed three versions of a delayed intention task with and without the use of reminders, along with three measures of working memory capacity. Results showed that individuals with higher working memory fulfilled more intentions when having to rely on their own memory, but this difference was eliminated when offloading was permitted. Individuals with lower working memory chose to offload more often, suggesting that they were less willing to engage in effortful maintenance of internal representations when given the option. Working memory was not associated with metacognitive confidence or optimal offloading choices based on point value. These findings suggest offloading may help circumvent capacity limitations associated with maintaining and remembering delayed intentions.

KEYWORDS:

• Prospective memory
• offloading
• reminders
• working memory
• individual differences

Acknowledgements

We thank Daisy Estrada, Marissa Kepple, and Sarah Monier for their assistance in data collection and scoring. Portions of the data were presented at the ARMADILLO Conference is San Antonio, TX.

All data is currently accessible: https://osf.io/fzdth/?view_only=4c1bebd518cf449e9fdd6fadf760c8d3

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 There were no significant differences in performance on Day 1 tasks between those participants that did or did not return for Day 2.

2 Gilbert et al. (2020) used point values ranging from 1–9, resulting in 9 choice trials. There were also 4 forced internal and 4 forced external trials. We used an abbreviated version of this task due to time constraints associated with completing a larger cognitive battery and completion of multiple versions of the offloading task. Values of 1 and 9 were removed because they produced the least amount of variation in offloading choices, resulting in only 7 choice trials. To keep the approximate proportions of choice and forced trials similar to Gilbert et al., we used 3 forced internal and 3 forced external trials.

3 Prior research also excluded participants with negative correlations between target value and the likelihood of choosing reminders, as this reflects a random or counter-rational choice strategy (Gilbert et al., 2020; Sachdeva & Gilbert, 2020). In the current study, averaged across all three tasks there were 25 participants that had a negative correlation between the two variables. Because the focus of the current study is on individual differences in offloading choices, we opted to retain these participants. Note that excluding these participants resulted in an identical pattern of results as reported below. There was also one participant that had equivalent average performance between internal and external trials that we did not exclude because this made no influence on the results.

4 The root mean square error of approximation (RMSEA) and the standardised root mean square residual (SRMR) both reflect the average squared deviation between the observed and reproduced covariances. In addition, the non-normed fit index (NNFI) and the comparative fit index (CFI), both of which compare the fit of the specified model to a baseline null model. NNFI, and CFI values greater than .90 and RMSEA and SRMR values less than .08 are indicative of acceptable fit.