Code for: Sleep Links Gist Abstraction to Veridical Memory

Sleep is important for memory consolidation. By strengthening the original memory trace, sleep may improve monitoring processes, and thus the distinction between veridical and false memories. Simultaneously, sleep may facilitate gist abstraction and thus enhance false memory generation. Here, this question is studied using the Deese–Roediger–McDermott (DRM) paradigm, where participants learn lists of semantically related words, constructed to lead to false retrieval of lure words (semantically linked to the lists, but never presented). Previous literature has found sleep to increase or decrease false memories, or no significant effect, which might be due to methodological variance, but can also be explained by low statistical power. In this large online study (sleep = 104, wake = 101, AM control = 99, PM control = 94), a preregistered replication of Diekelmann et al. (2010), we find no effect of sleep on false memories, nor on general memory performance. We also do not replicate the finding that sleep increases false memories when overall memory performance is low. Instead, we find that false memories increase after sleep when intrusion-adjusted memory performance is high. We interpret this as generalization process during sleep (gist), which helps to form veridical memory, but also generates false memories. README.txt: general instructions how to run the code; nagel_et_al_code.zip: analysis code and a local renv environment to create a local package library for reproducibility - contents of the .zip: - 01_exclusion.R, Checks which exclusion criteria (if any) apply to all participants who at least started part 1 of the study. .- 02_0_preprocessing.R, Auxiliary file run by all analysis scripts that e.g. creates subsets or aggregated versions of the data for analysis. - 02_1_analysis.R, Main analysis file: DRM performance analysis for the experimental (wake, sleep) and circadian control (AM control, PM control) groups. - 02_2_analysis_confidence.R, Analyses of confidence responses during the DRM paradigm, including the analysis of trial order effects reported in the Supplementary Materials. - 02_3_analysis_mak_et_al.R, Alternative DRM performance analysis, following the analysis strategy Mak et al. (2023) used (mainly reported in the Supplementary Materials). - 02_4_analysis_plots.R, Plots accompanying the analyses in 02_1_analysis.R. - 02_4_analysis_confidence_plots.R, Plots accompanying the analyses in 02_1_analysis_confidence.R. - 03_0_preprocessing_control.R, Auxiliary file run by all control measure analysis scripts that e.g. creates subsets or aggregated versions of the data for analysis. - 03_1_analysis_control.R, Analysis of control measures: Psychomotor Vigilance Task, Stanford Sleepiness and Verbal Fluency (including verbal fluency correlation plot) - 04_analysis_performance_intrusions.R, Analyses of how the interaction between group, memory performance and intrusions affects false memory generation (reported in the Supplementary Materials). Also includes plots shown in the Supplementary Materials., - 05_analysis_performance_intrusions_mak.R, Re-analysis of the data collected by Mak et al. (2023). Analyses of how the interaction between group, memory performance and intrusions affects false memory generation (reported in the Supplementary Materials). Also includes plots shown in the Supplementary Materials. peerReviewed acceptedVersion