A tutorial on regression-based norming of psychological tests with GAMLSS.

Area of Science:

• Psychometrics
• Statistical Modeling
• Psychological Assessment

Background:

• Norm-referenced scores are crucial for interpreting individual test performance within a reference population.
• Traditional norming methods can be inflexible, while regression-based norming offers greater adaptability and efficiency.
• Generalized additive models for location, scale, and shape (GAMLSS) are increasingly popular for deriving accurate norms.

Purpose of the Study:

• To introduce and explain the methodology of regression-based norming using GAMLSS.
• To provide a practical, step-by-step guide for applying GAMLSS in psychological test norming.
• To demonstrate the utility of GAMLSS with an example using the intelligence test IDS-2.

Main Methods:

• The tutorial outlines a 6-step process for designing normative studies and gathering data.
• It details selecting appropriate GAMLSS models for empirical scales and deriving normed scores.
• Methods include deriving scores for composite scales and visualizing results for scale property insights.

Main Results:

• Regression-based norming with GAMLSS provides flexible and potentially more realistic norms.
• The approach is efficient, potentially requiring smaller sample sizes for equivalent precision compared to traditional methods.
• The study illustrates the successful application of GAMLSS for norm estimation using real-world data.

Conclusions:

• Regression-based norming with GAMLSS is a powerful and practical approach for psychological test norming.
• The provided methodology and R code facilitate the implementation of GAMLSS for creating robust norms.
• This method enhances the interpretability and precision of psychological test scores.