Scripts and information units in future planning: Interactions between a past and a future planning task

Abstract

Research on future thinking has emphasized how episodic details from memories are combined to create future thoughts, but has not yet examined the role of semantic scripts. In this study, participants recalled how they planned a past camping trip in Australia (past planning task) and imagined how they would plan a future camping trip (future planning task), set either in a familiar (Australia) or an unfamiliar (Antarctica) context. Transcripts were segmented into information units that were coded according to semantic category (e.g., where, when, transport, material, actions). Results revealed a strong interaction between tasks and their presentation order. Starting with the past planning task constrained the future planning task when the context was familiar. Participants generated no new information when the future camping trip was set in Australia and completed second (after the past planning task). Conversely, starting with the future planning task facilitated the past planning task. Participants recalled more information units of their past plan when the past planning task was completed second (after the future planning task). These results shed new light on the role of scripts in past and future thinking and on how past and future thinking processes interact.

Keywords:

• Future thinking
• Planning
• Memory
• Scripts
• Mental time travel

Notes

¹We designed the paradigm in a way that would let us include a subtest at the end of the session replicating Klein et al. (2010)’s experiment. However, the condition of the testing and the sample size were relatively different from those in the original. As we did not replicate the results, we decided not to include them in this article. The data for the replication were collected after the data presented in this paper; therefore it cannot be considered as a confounding factor.

²One participant was in the “past planning then future planning in Antarctica” condition; the other was in the “future planning in Australia then past planning” condition. These two participants had a significant number of information units in a single category that inflated their total number of information units. For example, one participant simply listed every item of food he would pack. Including them in our analyses would not alter their outcome; on the contrary, it would increase our effect size and therefore gives an erroneous inflated view of our results.

³We used generalized eta squared to report effect sizes as this design has a within-subject variable (Bakeman, 2005; Lakens, 2013; Olejnik & Algina, 2003).

⁴Bootstrapping procedures are robust methods that can be used when some assumptions are violated (Field, 2009).