Supporting construct validity of the Evaluation of Daily Activity Questionnaire using Linear Logistic Test Models

Area of Science:

• Psychometrics
• Occupational Therapy

Background:

• Construct validity is crucial for measurement tools.
• Traditional methods often lack explanatory power regarding the underlying attribute-score relationship.
• Linear Logistic Test Models (LLTMs) offer a framework to decompose item difficulties based on theoretical properties.

Purpose of the Study:

• To support the construct validity of the Evaluation of Daily Activity Questionnaire (EDAQ).
• To utilize item properties within LLTMs to account for item difficulties.
• To provide a theoretical explanation for variations in item difficulties.

Main Methods:

• Dichotomized EDAQ responses were analyzed using Rasch models for item difficulty estimation.
• LLTMs were employed to predict item difficulties based on seven identified item properties.
• Item properties were rated by 39 Occupational Therapists, with weights derived using cumulative link mixed models.

Main Results:

• The Rasch model demonstrated acceptable fit and unidimensionality for EDAQ items.
• LLTM with error showed a significantly better model fit compared to the standard LLTM.
• A refined LLTM using significant properties explained 77.5% of the variance in estimated item difficulties.

Conclusions:

• The LLTM with error model provided a satisfactory theoretical explanation for differences in daily living task difficulty.
• This modeling approach supports the construct validity of the EDAQ.
• The study highlights the utility of LLTMs in elucidating the basis of measurement item difficulties.