Regression-based developmental models exemplified for Wisconsin Card Sorting Test parameters: statistics and software...

Area of Science:

• Neuroscience
• Developmental Psychology
• Statistics

Background:

• Accurate score prediction in clinical and developmental research relies on normative data and statistical methods.
• Traditional methods often use descriptive statistics like standard scores for age groups.

Purpose of the Study:

• To illustrate a multiple regression approach for predicting individual Wisconsin Card Sorting Test (WCST) scores.
• To model linear and curvilinear age effects on WCST variables in a developmental sample.
• To develop a statistics program for applying these regression models in applied research.

Main Methods:

• Utilized a dataset of 345 Wisconsin Card Sorting Test (WCST) scores from individuals aged 6 to 26 years.
• Applied multiple regression to model linear and curvilinear age effects for 11 WCST variables.
• Determined confidence limits for mean and individual score predictions.

Main Results:

• Curvilinear age effects significantly outperformed linear effects, indicating negatively accelerated developmental functions for WCST variables.
• Multiple regression models explained 2% to 26% of the variance in WCST scores.
• A statistics program was developed for individual score predictions using normative data and up to 7 predictors.

Conclusions:

• The multiple regression approach offers a robust method for individual score prediction in developmental research, surpassing conventional standard score methods.
• The findings highlight the importance of considering curvilinear age trajectories in cognitive development.
• The developed program facilitates the application of advanced statistical modeling in clinical and applied settings.