Dissociating indifferent, directional, and extreme responding in personality data: Applying the three-process model to self- and observer reports

Ingo Zettler, Jonas W B Lang, Ute R Hülsheger, Benjamin E Hilbig

    34 Citations (Scopus)

    Abstract

    Research suggests that respondents vary in their tendency to use the response scale of typical (Likert-style) questionnaires. We study the nature of the response process by applying a recently introduced item response theory modeling procedure, the three-process model, to data of self- and observer reports of personality traits. The three-process model captures indifferent, directional, and extreme responding. Substantively, we hypothesize that, and test whether, trait Honesty-Humility is negatively linked to extreme responding. We applied the three-process model to personality data of 577 dyads (self- and observer reports of the HEXACO Personality Inventory-Revised; Lee & Ashton,) of Dutch and German undergraduate respondents. First, we provide evidence that indifferent, directional, and extreme responding can be separated from each other in personality data through the use of the three-process model. Second, we show that the various response processes show a pattern of correlations across traits and rating sources which is in line with the idea that indifferent and extreme responding are person-specific tendencies, whereas directional responding is content-specific. Third, we report findings supporting the hypothesis that Honesty-Humility is negatively linked to extreme responding. In Likert-based personality data, applying the three-process model can unveil individual differences in the response process.

    Original languageEnglish
    JournalJournal of Personality
    Volume84
    Issue number4
    Pages (from-to)461-472
    ISSN0022-3506
    DOIs
    Publication statusPublished - 1 Aug 2016

    Fingerprint

    Dive into the research topics of 'Dissociating indifferent, directional, and extreme responding in personality data: Applying the three-process model to self- and observer reports'. Together they form a unique fingerprint.

    Cite this