Graded Minds: ERP Evidence for Continuous, Multidimensional Conceptual Processing in Picture–Word Stroop and Semantic Decision Tasks

This record contains the data, materials, and code accompanying the manuscript "Graded Minds: ERP Evidence for Continuous, Multidimensional Conceptual Processing in Picture–Word Stroop and Semantic Decision Tasks" (Di Pietro, Viviani, Ülkü, Vallesi, Montefinese, & Ambrosini; under review). The same materials are also available on the Open Science Framework (https://osf.io/4mzgj/). Study summary. EEG was recorded from 32 native Italian speakers performing a Picture-Word Stroop (PWS) task and a Semantic Decision Task (SDT) on the same stimulus set. Target-distractor similarity was quantified trial by trial along semantic (subjective ratings and FastText distributional vectors), vision-based (CNN-derived conceptual- and perceptual-visual similarity, ViSpa), and phonological dimensions. Behavioral and ERP data (stimulus- and response-locked) were analyzed with single-trial linear mixed-effects models (lmeEEG with TFCE). Stimulus-locked ERP analyses were run on the full sample (N = 32) and on the subsample with behavioral data (N = 21); behavioral and response-locked analyses are based on N = 21. Contents. Raw, post-ICA, and cleaned EEG data (64 channels, 500 Hz; BrainVision and EEGLAB formats), fully anonymized. Trial-level behavioral data and target-distractor similarity measures (subjective semantic, distributional-linguistic, conceptual-visual, perceptual-visual, phonological). Merged single-trial EEG matrices used for the lmeEEG analyses. Analysis code (MATLAB): EEG preprocessing, behavioral linear mixed-effects models, and mass-univariate ERP linear mixed-effects models with TFCE. Experimental materials (PsychoPy 2023.2.3): experiment files, trial lists, and picture stimuli with superimposed distractor words. How to use this record. The repository is split into several zip files. Download and extract all of them in the same main folder to reconstruct the folder structure (a_Behavioral_Analyses, b_EEG_Preprocessing, c_ERP_Analyses, Task). Then run the scripts in the order described in the README.md file, which also contains the full data dictionary, software requirements (MATLAB R2023a, EEGLAB 2024.0, lmeEEG), and the list of target-distractor pairs (Stimuli.txt). Notes. In data files and scripts the Picture-Word Stroop task is labeled "PWI"; it corresponds to the PWS task in the manuscript. Pictures derive from the Multilingual Picture Database (Multipic; Duñabeitia et al., 2022). Third-party tools (EEGLAB and plugins, lmeEEG, Multipic) keep their original licenses. Licenses. Data and materials: CC BY 4.0. Code: MIT.