Dataset DOI: 10.5061/dryad.mgqnk99br

Description of the Data and File Structure

This dataset accompanies the publication Processing in working memory boosts long-term memory representations and their retrieval (Sabo & Schneider, 2025). For detailed information on the experimental paradigm, data collection procedures, and EEG setup, please refer to the publication. In addition to EEG data, raw behavioral data are included and can be used independently for behavioral analyses.

Behavioral datasets are provided in the following compressed files:

• behav_Exp1.zip
• behav_Exp2.tar

For each participant, four associated files are provided:

1. XX_learning.xlsx – learning phase data
2. XX_WM.xlsx – working memory phase data
3. XX_retrieval.xlsx – retrieval phase data
4. XX_diary.xlsx – diary file with responses and response times across all phases

Note: Event-coded EEG files already incorporate the behavioral data.

Behavioral Data – Experiment 1 (behav_Exp1.zip)

Participants: VP002 – VP45

Files per participant:

• XX_learning.xlsx – Contains information about the stimuli presented during the learning phase of the experimental task.
  • trial – Indicates the trial number. Trial numbers within the learning phase range from 1–360.
  • imgNumber – Codes the number of the object image presented during the learning phase (1–240).
  • prePos – Codes the location of the object. There are four possible locations: 1 = top, 2 = right, 3 = bottom, 4 = left.
• XX_WM.xlsx – Contains information about the stimuli presented during the working memory phase of the experimental task.
  • trial – Indicates the trial number. Trial numbers within the working memory phase range from 1–240.
  • imgNumber_target – Codes the number of the target object presented during the task. Target objects were the tested items in the working memory task (1–240).
  • pos_target – Codes the location of the target object (1–4).
  • imgNumber_distractor – Codes the number of the distractor object presented in the memory array. Distractor objects were not tested in the working memory task (1–240).
  • pos_distractor – Codes the location of the distractor object (1–4).
  • imgNumber_new1 – Codes the identity of one of the filler objects in the memory array. Filler objects were scrambled versions of real images, each assigned a number code.
  • pos_new1 – Codes the location of imgNumber_new1 (1–4).
  • imgNumber_new2 – Codes the identity of the second filler object in the memory array (also a scrambled image with a number code).
  • pos_new2 – Codes the location of imgNumber_new2 (1–4).
  • imgNumber_probe - Codes the number of the test object presented at the end of the working memory trial.
  • pos_probe – Codes the position of the test object. Since the test object was always presented centrally, this value is always 0.
• XX_retr.xlsx – Contains information about the stimuli presented during the retrieval phase of the experimental task.
  • trial – Indicates the trial number. Trial numbers within the retrieval phase range from 1–360.
  • imgNumber – Codes the number of the object image presented during the retrieval phase (1–240).
  • presPos – Codes the presented position. Since all objects were centrally presented in this phase, the values are always 0.
  • pos_corr_answer – Codes the correct location that participants were required to report in this phase.
• XX_diary.xlsx – Contains behavioral data recorded throughout the entire experiment.
  • trial – Indicates the trial number. Trial numbers 1–360 correspond to the learning phase, 361–600 to the working memory phase, and 601–960 to the retrieval phase.
  • corr_pos – Indicates the correct location participants were required to report during each phase. In the learning phase, this was always 2, because participants had to click the right button to confirm the end of a trial. In the working memory phase, the correct response was either 1 or 3, coding “match” or “no match” respectively (the mapping was counterbalanced across participants). In the retrieval phase, responses could range from 1–4, allowing participants to report any of the four locations.
  • RT – Response times associated with each button press.
  • accuracy – Indicates whether the participant’s response was correct or incorrect (labelled as false). Accuracy values are missing during the learning phase, replaced with "---" because the button press was only used to proceed to the next trial.File: behav_Exp2.tar

Description: Excel files containing the behavioural data of every participant from Experiment 2.

Behavioral Data – Experiment 2 (behav_Exp2.tar)

Participants: VP01 – VP45

Files per participant:

• XX_learning.xlsx – Contains information about the stimuli presented during the learning phase of the experimental task.
  • trial – Indicates the trial number. Trial numbers within the learning phase range from 1–600.
  • imgNumber – Codes the number of the object image presented during the learning phase (1–300).
  • prePos – Codes the location of the object. There are four possible locations: 1 = top, 2 = right, 3 = bottom, 4 = left.
• XX_WM.xlsx – Contains information about the stimuli presented during the working memory phase of the experimental task.
  • trial – Indicates the trial number. Trial numbers within the working memory phase range from 1–480.
  • imgNumber_target – Codes the number of the target object presented during the task. Target objects were the tested items in the working memory task (1–300).
  • pos_target – Codes the location of the target object (1–4).
  • imgNumber_distractor – Codes the number of the distractor object presented in the memory array. Distractor objects were not tested in the working memory task (1–300).
  • pos_distractor – Codes the location of the distractor object (1–4).
  • imgNumber_new1 – Codes the identity of one of the filler objects in the memory array. Filler objects were scrambled versions of real images, each assigned a number code.
  • pos_new1 – Codes the location of imgNumber_new1 (1–4).
  • imgNumber_new2 – Codes the identity of the second filler object in the memory array (also a scrambled image with a number code).
  • pos_new2 – Codes the location of imgNumber_new2 (1–4).
  • imgNumber_probe - Codes the number of the test object presented at the end of the working memory trial.
  • pos_probe – Codes the position of the test object. Since the test object was always presented centrally, this value is always 0.
  • condition – Codes the type of retro-cue presented in the working memory task. Half of the cues were selective retro-cues and the other half were neutral retro-cues.
• XX_retr.xlsx – Contains information about the stimuli presented during the retrieval phase of the experimental task.
  • trial – Indicates the trial number. Trial numbers within the retrieval phase range from 1–600.
  • imgNumber – Codes the number of the object image presented during the retrieval phase (1–300).
  • presPos – Codes the presented position. Since all objects were centrally presented in this phase, the values are always 0.
  • pos_corr_answer – Codes the correct location that participants were required to report in this phase.
• XX_diary.xlsx – Contains behavioral data recorded throughout the entire experiment.
  • trial – Indicates the trial number. Trial numbers 1–600 correspond to the learning phase, 601–1080 to the working memory phase, and 1081–1681 to the retrieval phase.
  • corr_pos – Indicates the correct location participants were required to report during each phase. In the learning phase, this was always 2, because participants had to click the right button to confirm the end of a trial. In the working memory phase, the correct response was either 1 or 3, coding “match” or “no match” respectively (the mapping was counterbalanced across participants). In the retrieval phase, responses could range from 1–4, allowing participants to report any of the four locations.
  • RT – Response times associated with each button press.
  • Accuracy – Indicates whether the participant’s response was correct or incorrect (labelled as false). Accuracy values are missing during the learning phase, replaced with "---" because the button press was only used to proceed to the next trial.

EEG Data – Experiment 1 (EEG_coded_Exp1.zip)

Format: The compressed folder contains 86 files, half of which are .set files and the other half are .fdt files. This file type is EEGLAB specific (MATLAB structure). Participants are labelled VP002 – VP45, every participant has a corresponding .set and .fdt file. Because the EEG datasets are large, EEGLAB stores the raw/processed data in a separate binary file with the .fdt extension. The .set file contains a pointer to the .fdt file, so both must be present in the same directory to load the dataset correctly. Files are event-coded and ready for preprocessing (see scripts at: https://osf.io/ne9yt/).

EEGLAB fields:

Triggers in the EEG data have the following meaning:

• 254 = start of experiment
• 255 = end of experiment
• 100-123 = start of a block
• 200-223 = end of a block
• 1 =  object onset in the learning phase
• 2 = memory array onset in the working memory phase
• 3 = retro-cue onset in the working memory phase
• 4 = test object onset in the working memory phase
• 5 = object onset in the retrieval phase
• 241 = response 

EEG Data – Experiment 2 (EEG_coded_Exp2.tar)

Identical parameters and structure as Experiment 1 EEG dataset.

Code/software

The scripts used in the current project can be accessed on OSF: https://osf.io/ne9yt/

The scripts were written in MATLAB R2023b, and they use the EEGLAB toolbox. Please refer to the publication for the specific packages needed to run the scripts. 

Description of the data contained in DisciplineSpecificMetadata.json

The metadata file describes a human cognitive neuroscience dataset consisting of neural data acquired using electroencephalography (EEG). The data are in raw form, with no preprocessing applied. The experiment belonging to this dataset, focuses on long-term memory and is based on a combined episodic long-term memory and working memory task. In Experiment 1, participants first learned object–location associations for later recall. In the second phase, a retro-cue working memory paradigm was employed in which objects from the learning phase were presented alongside scrambled distractors; after a short delay, a 100 % informative cue indicated which object would be tested, and participants judged whether the cued object matched a centrally presented probe. Each object appeared four times in the same location during this phase. In the final retrieval phase, all learned objects were presented again, and participants indicated the remembered location using a four-button spatially aligned response device. Experiment 2 was identical to Experiment 1, except that the working memory task included also a neutral (non-infomative) cue, which did not offer any information about the to-be-tested object. Analyses focused on data from this long-term memory retrieval phase. The dataset includes data from 43 participants per experiment, and the full analysis pipeline is available at https://osf.io/ne9yt/. MATLAB was used for experimental control and JASP for Bayesian statistical analysis.

Human subjects data

All participants provided explicit informed consent for their data to be shared in de-identified form in a public repository. To ensure privacy and confidentiality, all data were de-identified prior to publication. Specifically, all personal identifiers (e.g., names, dates of birth, contact details) were removed. EEG data files were stripped of any embedded metadata that could link the data to individual participants. Participant codes were randomly assigned and do not contain any information that could enable re-identification. No video, audio, or other biometric data were included. These steps ensure that the data cannot reasonably be used to identify individual participants.