Exploring Aberrant Responses Using Person Fit and Person Response Functions.

Area of Science:

• Psychometrics
• Educational Measurement
• Item Response Theory

Background:

• Traditional person fit statistics often yield ambiguous interpretations of aberrant response behavior in testing.
• Understanding individual response patterns is crucial for accurate score interpretation and test validity.

Purpose of the Study:

• To investigate the utility of person response functions (PRF) in supplementing and clarifying person fit statistics.
• To determine if PRF can offer more absolute and detailed insights into person misfit compared to relative fit statistics.

Main Methods:

• A Rasch model analysis was conducted on data from 31 participants who employed guessing strategies on 63 multiple-choice items.
• Participants self-reported their guessing strategies after each item response.
• Item calibrations were anchored to established values for appropriate administration.

Main Results:

• Participants exhibited expected poor model-data fit.
• Person response functions (PRF) provided insights into absolute person fit, distinguishing it from the relative fit offered by traditional statistics.
• PRF identified specific item locations and methods contributing to person misfit.

Conclusions:

• Person response functions (PRF) enhance the interpretation of person fit by offering absolute model-fit information.
• PRF provide granular detail on how and where responses deviate from the model, aiding practitioners.
• This supplementary information assists in the appropriate use and interpretation of individual test scores, especially when guessing strategies are employed.