Evaluating Gaze Event Detection Algorithms - Impacts on Machine Learning-based Classification and Psycholinguistic Statistical Modeling

Sep 17, 2026

Uploaded by Theodora

Eye movements provide non-invasive evidence about cognitive processing and are used in both psycholinguistic research and machine-learning pipelines, where fixations and saccades depend on gaze event detection algorithms. This work evaluates two threshold-based methods, I-DT and I-VT, and tests how algorithm choice affects downstream machine-learning classification and psycholinguistic statistical modeling. Across multiple eyetracking datasets and sampling frequencies, results show significant performance differences, with I-DT generally outperforming I-VT, and psycholinguistic findings aligning only when thresholds match established fixation metrics.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages16

Reading Time40 min

AccessFree to download

Introduction

Fixations and saccades detection

Study goals and evaluated algorithms

Zurich Open Repository and Archive University of Zurich University Library Strickhofstrasse 39 CH-8057 Zurich [www.zora.uzh.ch](www.zora.uzh.ch) Year: 2025 Evaluating Gaze Event Detection Algorithms: Impacts on Machine Learning-based Classification and Psycholinguistic Statistical Modeling Reich, David Robert ; Prasse, Paul ; Jäger, Lena A DOI: [https://doi.org/10.1145/3725835](https://doi.org/10.1145/3725835) Posted at the Zurich Open Repository and Archive, University of Zurich ZORA URL: [https://doi.org/10.5167/uzh-278292](https://doi.org/10.5167/uzh-278292) Journal Article Published Version The following work is licensed under a Creative Commons: Attribution 4.0 International (CC BY 4.0) License. Originally published at: Reich, David Robert; Prasse, Paul; Jäger, Lena A (2025) . Evaluating Gaze Event Detection Algorithms: Impactson Machine Learning-based Classification and Psycholinguistic Statistical Modeling. Proceedings of the ACM on Human-Computer Interaction, 9(3):ETRA10 . DOI: [https://doi.org/10.1145/3725835](https://doi.org/10.1145/3725835) Evaluating Gaze Event Detection Algorithms: Impacts on Machine Learning-based Classification and Psycholinguistic Statistical Modeling DAVID R. REICH∗ , University of Zurich, Switzerland; and University of Potsdam, Germany PAUL PRASSE∗ , University of Potsdam, Germany LENA A. JÄGER, University of Zurich, Switzerland Eye movements offer valuable, non-invasive insights into cognitive processes and are widely used in both psycholinguistic research and machine-learning applications, such as assessing reading comprehension and cognitive load. These applications typically rely on fixations and saccades detected through gaze event algorithms, which may be either proprietary or open-source. The impact of different gaze event detection algorithms on subsequent analysis is underexplored and often overlooked. This study investigates how two threshold-based algorithms, I-DT and I-VT, influence both machine-learning classification tasks and psycholinguistic statistical modeling. Using diverse datasetsÐincluding stationary, remote, and VR eyetracking data across multiple sampling frequenciesÐour findings show significant differences in downstream performance. For ML tasks, I-DT generally outperforms I-VT, with I-VT being highly sensitive to threshold choices. In psycholinguistic analysis, results confirm established findings only when thresholds align with established fixation metrics, emphasizing the importance of appropriate threshold selection for meaningful analysis. Our code is publicly available: [https://github.com/aeye-lab/eye-movement-preprocessing](https://github.com/aeye-lab/eye-movement-preprocessing). CCS Concepts: · Computing methodologies → Machine learning; Cross-validation; Classification and regression trees; Batch learning; Supervised learning by classification. Additional Key Words and Phrases: psycholinguistic analysis, psycholinguistic statistical modeling, machine learning, eye movements, I-DT, I-VT ACM Reference Format: David R. Reich, Paul Prasse, and Lena A. Jäger. 2025. Evaluating Gaze Event Detection Algorithms: Impacts on Machine Learning-based Classification and Psycholinguistic Statistical Modeling. Proc. ACM Hum.-Comput. Interact. 9, 3, Article ETRA10 (May 2025), 15 pages. [https://doi.org/10.1145/3725835](https://doi.org/10.1145/3725835) 1 Introduction Eye-tracking data has become an invaluable resource for understanding human cognitive processes across disciplines, including human-computer interaction, psychology, and cognitive science [15, 17, 22] . Through eye movements, researchers gain insights into human behavioral processes such as decision-making, visual search, and reading comprehension [1] . Key to these analyses is the accurate detection of gaze eventsÐspecifically, fixations during which the eye is relatively stable and information is processed, and saccades, rapid relocation movements between fixations

FAQs

Which two gaze event detection algorithms are compared in the study?

The study compares two threshold-based algorithms: I-DT and I-VT. Their effects are evaluated across downstream tasks.

How do the algorithms influence machine-learning classification performance?

For machine-learning tasks, I-DT generally outperforms I-VT. The results also indicate that I-VT is highly sensitive to threshold choices.

What determines whether psycholinguistic statistical modeling results align with established findings?

Psycholinguistic analysis confirms established findings only when detection thresholds align with established fixation metrics, highlighting the importance of correct threshold selection.

Sep 17, 2026

Uploaded by Theodora