LCYA: Concurrent Effort in Younger Adults

Overview

This repository contains the analysis code and supplementary materials for:

Dastgheib, M.*, Sun, A. Y.*, Yaghoubi, K. C., Zhang, W., Peters, M. A. K., Bennett, I. J., & Seitz, A. R. (in press). Effects of concurrent task difficulty and physical effort on memory and perceptual performance and pupil response in younger adults. *co-first authors

The study examined effects of concurrent task difficulty (Easy vs. Hard stimulus similarity) and physical effort (Low 5% vs. High 40% MVC handgrip) on behavioral performance and pupillometric responses across three cognitive domains—visual working memory, auditory perception, and visual perception—in healthy younger adults (N = 38, M age = 22.9 years).

Key findings:

• Robust main effects of task difficulty on accuracy (all three tasks) and reaction time (perceptual tasks)
• Main effect of physical effort on reaction time for the visual perception task only
• Main effect of physical effort on pupillary response (Total AUC) for all three tasks
• Exploratory main effect of task difficulty on Cognitive AUC for the auditory perception task
• No interactions between task difficulty and physical effort for any outcome
• No significant individual-difference correlations between behavioral and pupillary effort effects

These findings suggest that task difficulty and physical effort influence performance and pupillary arousal through largely independent mechanisms in healthy younger adults.

Study Design

Tasks

Participants completed three experimental sessions, each involving a different cognitive task performed concurrently with an isometric handgrip at either Low (5% MVC) or High (40% MVC) force:

• CDT (Change Detection Task): Visual working memory — participants judged whether a probe arrow orientation matched a studied arrow. Task difficulty was manipulated by varying the rotation difference between studied and probe arrows (Easy: 45° or 90°; Hard: 5° or 20°).
• ADT (Auditory Discrimination Task): Auditory frequency discrimination — participants judged whether two sequential tones were the same or different. Task difficulty was manipulated by varying the frequency offset (Easy: +32 or +128 Hz; Hard: +4 or +8 Hz).
• VDT (Visual Discrimination Task): Visual contrast discrimination — participants judged whether two sequential Gabor patches were the same or different. Task difficulty was manipulated by varying the contrast offset (Easy: +0.16 or +0.32; Hard: +0.04 or +0.08).

Physiological Measure

Pupil diameter was recorded continuously at 1000 Hz (EyeLink 1000, SR Research) and downsampled to 100 Hz. Two AUC metrics were computed:

• Total AUC: Integrated baseline-corrected pupil response from squeeze onset to response window onset
• Cognitive AUC (exploratory, ADT/VDT only): Integrated pupil response from 300 ms after target onset until response window onset, corrected to a pre-target baseline

Repository Structure

LCYA/
├── analysis/
│   ├── main/                        # Full publication pipeline (R Markdown)
│   ├── manuscript_scripts/          # Scripts for individual analyses and figures
│   │   ├── figures/                 # Figure generation scripts (Figures 2–4)
│   │   └── *.R                      # Statistical analysis scripts
│   └── revision_analyses/           # Supplementary analyses added at revision
│       ├── *.R                      # Analysis scripts (numbered 01–05)
│       └── outputs/                 # Generated tables and figures
├── code/
│   ├── plot_objects/                # ggplot theme and task-specific plot templates
│   ├── preprocessing/               # Data preprocessing (MATLAB scripts, Python notebooks)
│   └── utilities/                   # Shared helper functions
├── documentation/
│   └── results_reports/             # Statistical summaries and reports
└── figures/
    └── publication/                 # Final publication-ready figures (PNG, 300 DPI)

Key Analysis Scripts

Primary Analyses

Analysis	Script	Description
Full pipeline	analysis/main/Aggregated_analysis_publication.Rmd	Complete R Markdown pipeline reproducing all results
Behavioral + AUC models	analysis/manuscript_scripts/LCYA_Final_Corrected_Analysis.R	Primary GLMMs and LMMs for accuracy, RT, and pupil AUC
AUC models	analysis/manuscript_scripts/LCYA_AUC_Analysis.R	Focused LMM analysis for Total and Cognitive AUC
Individual differences	analysis/manuscript_scripts/Bivariate_Individual_Differences_brms.R	Bayesian multivariate models for behavior–pupil correlations

Manuscript Figures

Figure	Script	Description	Output
Figure 2 – Behavioral performance	analysis/revision_analyses/LCYA_Updated_Figures_Individual_Points.R	Accuracy and RT by task difficulty × physical effort for CDT, ADT, VDT, with individual participant means and connecting lines	figures/publication/Figure2_Task_Performance_Effects_Clean_Final.png
Figure 3 – Pupil waveforms	analysis/manuscript_scripts/figures/LCYA_Isolated_Cognitive_AUC_Analysis.R	GAM-smoothed pupil waveforms with pre-trial baseline and AUC windows annotated; CIs derived from subject-level means	figures/publication/Figure3_Pupil_Waveforms.png
Figure 4 – AUC bar plots	analysis/revision_analyses/LCYA_Merged_TotalCognitive_AUC_IndivPoints.R	Merged Total AUC (all tasks) and Cognitive AUC (ADT/VDT) with individual data points	figures/publication/Figure4_Merged_Total_Cognitive_AUC.png

Figure 3 details:

• Baseline: 500 ms pre-squeeze window (B₀)
• CDT: Total AUC only (Cognitive AUC excluded — probe presented during response window)
• ADT/VDT: Total AUC + Cognitive AUC shown; pre-stimulus baseline (B₂b) annotated

Revision Analyses (analysis/revision_analyses/)

All scripts output tables to analysis/revision_analyses/outputs/.

Script	Supplement	Description
LCYA_Random_Slopes_Models.R	Section 2	Trial-level GLMMs (accuracy) and LMMs (RT, Total AUC, Cognitive AUC) with maximal random-effects structure (Barr et al., 2013)
LCYA_TOST_Equivalence_Testing.R	Section 3	Two one-sided equivalence tests (TOST) for all non-significant interaction and PE main effects; SESOI anchored to Park et al. (2021), d = 0.79
LCYA_Grip_Force_Continuous_Models.R	Section 4	Models replacing binary Physical Effort with actual measured grip force (AUC/MVC, z-scored); includes motor-artifact check in High-effort trials
LCYA_Confidence_Ratings_Analysis.R	Section 1	Trial-level LMMs for subjective confidence ratings (1–4 scale) across all tasks
LCYA_Merged_TotalCognitive_AUC_IndivPoints.R	Figure 4	Generates Figure 4 (merged Total + Cognitive AUC bar plots with individual points)
LCYA_Updated_Figures_Individual_Points.R	Figure 2	Generates Figure 2 (behavioral performance with individual data points)

Key output files:

• LCYA_FixedEffects_AllModels.csv — fixed-effect estimates (b, SE, 95% CI, p) for all primary models (Supplementary Table S6)
• LCYA_GripForce_FixedEffects.csv — estimates from continuous grip force models (Supplementary Tables S4–S5)
• LCYA_TOST_Results.csv + LCYA_TOST_ForestPlot.png — equivalence test results (Supplementary Table S3 + Figure S2)
• FigureS_ActualForce_vs_TotalAUC.png — dose-response figure (Supplementary Figure S3)

Dependencies

R Packages

install.packages(c(
  "tidyverse",   # data wrangling + ggplot2
  "lme4",        # mixed-effects models
  "lmerTest",    # p-values for LMMs
  "mgcv",        # GAM smoothing (Figure 3)
  "patchwork",   # combining plots
  "cowplot",     # plot assembly
  "viridis",     # color palettes
  "brms"         # Bayesian individual-differences models
))

R version used: 4.5.1

Data Preprocessing

• MATLAB (code/preprocessing/): eyetracker file conversion
• Python notebooks (code/preprocessing/python_notebooks/): trial-level data aggregation per task

See code/preprocessing/README.md for preprocessing details.

Usage

Reproducing Manuscript Figures

1. Install all required R packages (see Dependencies)
2. Set BASE_DIR at the top of each script to point to your local data directory:

   BASE_DIR <- "/path/to/your/LC-YA"

3. Run in order:

   source("analysis/revision_analyses/LCYA_Updated_Figures_Individual_Points.R")
   source("analysis/manuscript_scripts/figures/LCYA_Isolated_Cognitive_AUC_Analysis.R")
   source("analysis/revision_analyses/LCYA_Merged_TotalCognitive_AUC_IndivPoints.R")

Full Analysis Pipeline

rmarkdown::render("analysis/main/Aggregated_analysis_publication.Rmd")

Data Requirements

Raw data are not included (see Data Availability below). Scripts expect:

• Pupil data: 100 Hz/ directory containing *_DS100_merged.csv files with columns sub, trial_index, time, pupil, duration_label, stimLev, isStrength
• Behavioral data: Complete_Manuscript_Results/complete_analysis_data.csv

Data Availability

De-identified behavioral and pupillometry data are available from the corresponding author upon reasonable request (subject to IRB/consent constraints). Analysis code and study materials are publicly available via Zenodo: https://doi.org/10.5281/zenodo.18204999

Citation

@article{dastgheib2025concurrent,
  title   = {Effects of concurrent task difficulty and physical effort on memory
             and perceptual performance and pupil response in younger adults},
  author  = {Dastgheib, Mohammad and Sun, Andrew Y. and Yaghoubi, Kimia C. and
             Zhang, Weiwei and Peters, Megan A. K. and Bennett, Ilana J. and
             Seitz, Aaron R.},
  year    = {2025},
  note    = {in press}
}

Contact

Mohammad Dastgheib — mdast003@ucr.edu
University of California, Riverside

License

This project is licensed under the GPLv3 License — see the LICENSE file for details.