Episodic future thinking: Linking neuropsychological performance with episodic detail in young and old adults

Abstract

Episodic future thinking (EFT) has been linked with our ability to remember past events. However, its specific neurocognitive subprocesses have remained elusive. In Experiment 1, a study of healthy older adults was conducted to investigate the candidate subprocesses of EFT. Participants completed a standard EFT cue word task, two memory measures (Verbal Paired Associates I, Source Memory), and two measures of executive function (Trail Making Test, Tower Test). In Experiment 2, healthy young adults also completed an EFT task and neuropsychological measures. The link between neurocognitive measures and five characteristics of EFT was investigated. Specifically, it was found that Source Memory and Trail Making Test performance predicted the episodic specificity of future events in older but not younger adults. Replicating previous findings, older adults produced future events with greater semantic but fewer episodic details than did young adults. These results extend the data and emphasize the importance of the multiple subprocesses underlying EFT.

Keywords:

• Episodic future thinking
• Source memory
• Executive function
• Episodic memory
• Imagination

Acknowledgments

This study was conducted as part of a PhD programme, which was made possible with financial support from the Institute of Psychological Sciences, University of Leeds, to which the first author extends his gratitude. The authors would also like to thank Katalin Pauly-Takacs and Lucy Justice for being second coders in this study.

Notes

¹These subprocesses may be recruited differentially depending on the cognitive function employed (see Hassabis & Maguire, 2007).

² A task that has been used as an index of episodic memory (see Paraskevaides et al., 2010).

³McDonough and Gallo (2010) have explored the role of reality monitoring in episodic future thinking, but focused upon the subsequent source memory for previously constructed memories and future events. Although it was important in emphasizing the differences between past and future event constructions, it is not relevant to the current study, which used source memory as an individual difference measure.

⁴ However, it should be noted that some of these differences may be due to differences in narrative style, as shown in a recent study by Gaesser and colleagues (2011) who demonstrated similar young–old differences using the Levine coding scheme when participants described pictures of scenes as well as when remembering/imagining autobiographical events.

⁵ First eight words taken from Dalla Barba, Cappelletti, Signorini, and Denes (1997).

⁶Randomized lists were generated using an online randomizer tool (Urbaniak & Plous, 2007).

⁷ It was unfeasible to code temporal distance (as in Anderson et al., 2012) due to a paucity of temporal information across future events—this was the case for young and old adults.

⁸Recognition (but not source memory) was inversely related to age (r = –.42, p < .05), which may demonstrate the difficulty of this recognition task. The lack of an age relationship for source memory may be due to the fact that those who recognized fewer items made fewer source memory judgements.

⁹Spatial detail, as measured by LIWC (Pennebaker et al., 2007), correlated positively with spatial reference developed by Hassabis et al. (2007), r = .55, p < .005, based upon a sample of 28 future events.

¹⁰Tower Test Performance was not correlated with the Speed or Switching condition of the Trails test.

¹¹The authors recognize that the lack of statistical power increased the chance of a Type I error. However, this would not explain why some measures correlated with episodic specificity while some did not.

¹² The full young adult study consisted of two other conditions, which involved dual-task paradigms to assess the effect of a secondary task upon future thinking. However, in contrast to Anderson, Dewhurst, and Nash (2012), no effects of the secondary task were found on future event characteristics. With the aim of contributing the most salient findings to the field, only the standard future thinking condition from this study is reported here to extend findings from Experiment 1 and to extend the evidence base of age-related effects on the ability to simulate episodic future events. However, we are developing research on future thinking using the dual-task paradigm and hope to develop a more effective experimental design.

¹³We thank an anonymous reviewer for bringing this to our attention.