Maximum information approach to scale description for affective measures based on the Rasch model.

Area of Science:

• Psychometrics
• Educational Measurement
• Social Sciences

Background:

• Affective measures like attitude scales are crucial for understanding respondent sentiments.
• Qualitative interpretation of such data often presents challenges in psychometric analysis.
• Existing methods may not fully capture the nuanced meaning within response categories.

Purpose of the Study:

• To develop a novel method for the qualitative interpretation of data from affective measures.
• To enhance the understanding of response category meanings within psychometric models.
• To provide a framework for linking latent trait locations to observable respondent behaviors.

Main Methods:

• Utilizing the Rasch model for ordered categories to analyze affective scale data.
• Applying the Bock procedure to partition item information across response categories.
• Determining response category locations by maximizing associated information.
• Clustering response categories around latent trait points for qualitative description.

Main Results:

• Successfully demonstrated a method to qualitatively interpret affective data.
• Identified specific locations for response categories on the latent trait continuum.
• Provided explicit behavioral descriptions corresponding to different points on the latent trait.
• Illustrated the method's application using an alumni survey dataset.

Conclusions:

• The proposed method offers a robust approach to qualitative interpretation of affective measures.
• This technique enhances the utility of psychometric instruments by providing meaningful descriptions of response categories.
• Findings facilitate a deeper understanding of respondent attitudes and behaviors in survey research.