This repository contains the data and the R script used to analyze the data for the article titled "The DRM illusion in short-term memory: Opposite effects of retention interval on true and false recognition." This submission includes: The raw data file: source_data_UTF8.csv The R script for analysis: R_script_with_comments.R or script_UTF8.txt A file with a description of each variable: variable description.pdf or variable_description_UTF8.txt The DRM lists used: DRM list.pdf Below is the summary of the article: A short-term memory (STM) version of the Deese/Roediger-McDermott (DRM) paradigm was employed to investigate how true and false recognition evolved as STM contents were lost over a short time window immediately after initial encoding. Presentations of five-word DRM lists were followed by list-specific recognition tests applied either immediately or after a distractor-filled retention interval of 3, 9, or 27 s. Results showed a decrease in the probability of true recognition and an increase in the probability of false recognition as the retention interval grew longer. Based on the fuzzy-trace theory, we suggest that this pattern emerged from the different durability of item-specific phonological representations, which would play the dual role of supporting true memory and preventing false recognition, and integrative semantic representations, whose overlap with the critical items would give rise to the DRM illusion. As a further contribution, our study helps establish the DRM illusion in STM as a genuine and robust phenomenon by showing that it can be observed even with no delay and no intervening distraction between study and test, and that it occurs when participants are explicitly warned about the presence of semantically related distractors and instructed not to be misled by them. Keywords: False memory, DRM, short-term memory/working memory, recognition, retention interval, fuzzy-trace theory, explicit warnings, Bayesian multilevel modeling.