Simple regression-based procedures for taxometric investigations

Area of Science:

• Psychopathology
• Quantitative Psychology
• Statistical Modeling

Background:

• Theories of psychopathology often propose distinct latent populations, termed 'taxa'.
• Existing taxometric procedures for identifying these taxa are statistically robust but algebraically complex and lack accessible software.
• These limitations hinder adoption by non-statistician researchers.

Purpose of the Study:

• To present a simplified taxometric procedure, MAXSLOPE, to address the accessibility issues of current methods.
• To provide an intuitive, graphically based approach for identifying latent taxa.
• To demonstrate the utility of MAXSLOPE with artificial datasets, including non-ideal cases.

Main Methods:

• Developed the MAXSLOPE procedure, a taxometric method based on regression plots of variable pairs.
• Implemented MAXSLOPE using commonly available statistical software for ease of use.
• Applied MAXSLOPE to two artificial datasets to illustrate its application and effectiveness.

Main Results:

• The MAXSLOPE procedure successfully identified clear-cut evidence of taxa in one artificial dataset.
• The graphical approach demonstrated efficacy even in a second dataset with less distinct grouping structures.
• Comparisons were made with traditional cluster analysis and multivariate normal mixture analysis.

Conclusions:

• MAXSLOPE offers a more intuitive and accessible taxometric procedure for psychopathology research.
• The graphical method simplifies the identification of latent populations, overcoming prior algebraic and software barriers.
• This approach enhances the practical application of taxometric principles in understanding individual differences.