Eye Fixation Allocation in Divided Attention Task

Four adult participants completed a delayed matching-to-sample task under five reinforcement ratio conditions. The dataset contains anonymized, processed eye-tracking data including session-level fixation allocation, trial-level saccade data, and fixation-by-fixation reports. It was created by João Lucas Bernardy and colleagues and deposited in Harvard Dataverse in 2026.

Use Cases

• Model the ratio of total fixation duration on left vs. right sample elements against reinforcement ratios to test the generalized matching law.
• Analyze exit latency measures from trial-level saccade data to understand reaction time as a function of reinforcement uncertainty.
• Use fixation-by-fixation and saccade-by-saccade reports to study the microstructure of attentional allocation in a divided attention paradigm.
• Reproduce statistical models and figures from the manuscript using the provided R script and aggregated session-level data.

Strengths

• Data from a controlled within-subject design with five systematically manipulated reinforcement ratios (1:9 to 9:1).
• Includes raw fixation-level and saccade-level eye-tracking records alongside aggregated analysis files.
• Accompanied by an R script for full reproducibility of the reported analyses.

Limitations

• Small sample size of only four participants, limiting statistical power and generalizability.
• Geographic and demographic coverage is unspecified, likely representing a convenience sample.

Provenance

Source

Harvard Dataverse, from the study 'Extending the Matching Law to the Allocation of Eye Fixations in a Divided Attention Task'.

Collection Method

Eye movements were recorded with an EyeLink 1000 Plus eye tracker during a delayed matching-to-sample task with compound samples.

Freshness

Last updated in March 2026.