The Graded Response Differential Discrimination Model Accounting for Extreme Response Style

Area of Science:

• Psychometrics
• Behavioral Science

Background:

• Individual differences in response spacing, such as extreme response style, introduce bias in questionnaire data analysis.
• The differential discrimination model was recently proposed to adjust for response bias in continuous data.

Purpose of the Study:

• To extend the differential discrimination model for analyzing graded response formats common in behavioral sciences.
• To provide a method for accounting for response bias in ordinal questionnaire data.

Main Methods:

• Extension of the differential discrimination model to accommodate graded response items.
• Application of the extended model using standard structural equation modeling software.
• Empirical examples and a simulation study to validate the approach.

Main Results:

• The extended differential discrimination model effectively analyzes graded response data.
• The model successfully adjusts for response bias in ordinal questionnaire formats.
• Structural equation modeling provides a feasible framework for fitting the extended model.

Conclusions:

• The extended differential discrimination model is a significant advancement for analyzing questionnaire data with graded response formats.
• This approach enhances the accuracy of psychometric analyses in behavioral sciences by accounting for response biases.
• Researchers can now apply advanced bias correction techniques to a wider range of commonly used questionnaire types.