About this repository

This data repository contains datasets from an eye-tracking study involving 121 participants who completed tasks on six different websites. Each participant was assigned a specific task and given up to three minutes to locate the required information. After completing the task, participants were allowed to freely browse the website for the remainder of the session. A more detailed description of the dataset can be found in the publication linked under Related Works.

The present data corresponds to the analysis performed in the publication "Gaze Behavior in task- and non-task-related Web Interaction: Implications for Usability and UX," published as part of the 2nd Workshop on Eye Tracking Techniques, Applications and Challenges (ETTAC), held at the International Conference on Pattern Recognition (ICPR) 2026 in Lyon, France. The research goal of the associated paper was to analyze how task-related and non-task-related interaction differ in terms of basic eye movement metrics, including fixation duration, fixation coordinates, saccade duration, and saccade length. The analysis examined both overall differences between the two interaction states and 20-second time windows surrounding task completion, to investigate temporal changes associated with participants' transitions between interaction modes. 

Abstract

Several studies show that eye movements differ between task-related search and non-task-related free web browsing. Yet, experiments typically rely on simplified single-page websites or screenshots. This paper extends that line of research by analyzing gaze recordings from 121 participants interacting with multi-page websites. We investigated whether core eye-tracking metrics systematically differ between both user states in complex real-world web interactions. The analysis revealed statistically significant effects for fixation duration, fixation distribution (x- and y-coordinates), and saccade duration, with medium to large Cohen’s d effect sizes. Consequently, our findings highlight that task- and non-task-related user states persist in real-world web contexts and represent distinct interaction modes characterized by systematically different gaze patterns. This distinguishability has important implications for the design of usability- and UX-oriented eye-tracking studies, as both product qualities are defined by task-relatedness. Our results suggest that certain eye movements may be more usability- or UX-related depending on the user’s interaction state.

Data Explanation

This repository includes two main files:

1. the raw dataset, which contains the eye movements (fixations and saccades) each participant performed on the respective stimuli
2. a dataset containing the timestamps related to task completion

• Timestamp Dataset
  The dataset contains the following columns:

  • Participant: Identifier of the participant (e.g., P001)

  • Stimulus: Indicates the website associated with the respective eye movements. Possible values include DAV_1, DAV_2, DAV_3, WATER_1, WATER_2, and WATER_3, referring to three websites each from the German Alpine Association (DAV) and German water providers.

  • Eye movement id: Identifier associated with the eye movements of a participant. Every interaction starts with the ID 1 and counts up

  • Eye movement type: Indicates whether an eye movement is a fixation, a saccade or unknown. Classification is based on Tobii I-VT filter. Any data gaps are subsequently imputed if all of the following conditions are met: (1) the eye movements preceding and following the gap are of the same type, (2) the gap duration does not exceed 200 ms, and (3) the resulting imputed eye movement duration does not exceed the mean ± 3 SD of eye movement durations for the given participant–stimulus combination.
    
    • Flag = 0: No missing data
    • Flag = 1: Fixation located outside the screen boundaries (±1° tolerance)
    • Flag = 2: Missing data occurred between two eye movements of different types
    • Flag = 3: Missing data gap exceeds 200 ms
    • Flag = 4: Missing data occurred between two fixations, but the spatial displacement between the endpoint of the first fixation and the start point of the second fixation exceeds 1°
    • Flag = 5: The duration of the imputed eye movement exceeds the mean ± 3 SD for the given participant–stimulus combination.
    • Flag = 9: Assigned to eye movements adjacent to a non-imputable gaze sequence.Eye Movement Dataset
      The dataset contains the following columns:

  • Start timestamp: A Long value based on the Tobii eye tracker timestamps indicating the start of an eye movement.

  • End timestamp: A Long value based on the Tobii eye tracker timestamps indicating the end of an eye movement.

  • Fixation point X: Represents the x-coordinate of a fixation. This value is left blank for saccades or unknown eye movements.

  • Fixation point Y: Represents the y-coordinate of a fixation. This value is left blank for saccades or unknown eye movements.

  • Duration: Duration of an eye movement in ms. It is calculated by subtracting the start timestamp of an eye movement from the end timestamp.

• Timestamp Dataset
  The dataset contains the following columns:

  • Participant: Identifier of the participant (e.g., P001)

  • Stimulus: Indicates the website associated with the respective eye movements. Possible values include DAV_1, DAV_2, DAV_3, WATER_1, WATER_2, and WATER_3, referring to three websites each from the German Alpine Association (DAV) and German water providers.

  • Reported difficulty: An Integer value ranging from 1 to 10 capturing the self-reported difficulty of a given task

  • First right click timestamp: A Long value based on the Tobii eye tracker timestamps associated with the first right click of a participant. A value of 0 indicates that the task was not completed, while a value of #NV is used in case of ambiguous inputs.

  • Second right click timestamp: A Long value based on the Tobii eye tracker timestamps associated with the second right click of a participant. A value of 0 indicates that the task was not completed, while a value of #NV is used in case of ambiguous inputs.

  • Time between right clicks: Time elapsed between both right clicks represented in seconds.

Acknowledgments

This study was conducted as part of the EU-funded project "Digital Innovation Ostbayern (DInO)". DInO is part of the European Digital Innovation Hub (EDIH) program and is funded by

• the European Union (Project Reference 101083427) and
• the European Funds for Regional Development (EFRE) (Project Reference 20-3092.10-THD-105).

This eye-tracking study was approved by the Joint Ethics Committee of the Bavarian Universities (GEHBa) with the reference number GEHBa-202312-V-155-R.

The Tobii Pro Fusion eye-trackers were funded by the ‘German Federal Ministry of Research, Technology and Space’ (BMFTR) through the granting of the 'Bavarian State Budget' (ZD.B) (FKZ: 16-1541).

Contact Information

If you have any questions feel free to reach out to the owner of this repository by mail: fabian.engl(a)oth-regensburg.de