An exploratory study of eye typing fundamentals: Dwell time, text entry rate, errors, and workload

Kari Jouko Räihä, Saila Ovaska

    Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review

    38 Citations (Scopus)

    Abstract

    Although eye typing (typing on an on-screen keyboard via one's eyes as they are tracked by an eye tracker) has been studied for more than three decades now, we still know relatively little about it from the users' point of view. Standard metrics such as words per minute and keystrokes per character yield information only about the effectiveness of the technology and the interaction techniques developed for eye typing. We conducted an extensive study with almost five hours of eye typing per participant and report on extended qualitative and quantitative analysis of the relationship of dwell time, text entry rate, errors made, and workload experienced by the participants. The analysis method is comprehensive and stresses the need to consider different metrics in unison. The results highlight the importance of catering for individual differences and lead to suggestions for improvements in the interface.

    Original languageEnglish
    Title of host publicationConference Proceedings - The 30th ACM Conference on Human Factors in Computing Systems, CHI 2012
    Pages3001-3010
    Number of pages10
    DOIs
    Publication statusPublished - 2012
    Publication typeA4 Article in conference proceedings
    Event30th ACM Conference on Human Factors in Computing Systems, CHI 2012 -
    Duration: 1 Jan 2012 → …

    Conference

    Conference30th ACM Conference on Human Factors in Computing Systems, CHI 2012
    Period1/01/12 → …

    Keywords

    • Adjustable dwell time
    • Error analysis
    • Extended study
    • Eye tracking
    • Eye typing
    • Workload

    Publication forum classification

    • Publication forum level 1

    Fingerprint

    Dive into the research topics of 'An exploratory study of eye typing fundamentals: Dwell time, text entry rate, errors, and workload'. Together they form a unique fingerprint.

    Cite this