A Multidimensional Model to Facilitate Within Person Comparison of Attributes

Area of Science:

• Psychometrics
• Psychological Measurement
• Quantitative Psychology

Background:

• Within-person comparisons in psychology often assume equal units (EU) and/or equal origins.
• This assumption of measurement validity is rarely tested or validated.
• Existing methods lack robust validation for comparing psychological traits.

Purpose of the Study:

• To present and validate a multidimensional structural equation modeling/item response theory (SEM/IRT) model.
• To demonstrate how conjoint measurement principles can establish equal units (EU) for psychological scales.
• To improve the validity of within-person comparisons in psychological research.

Main Methods:

• Described a multidimensional SEM/IRT model from existing literature.
• Applied principles of conjoint measurement to assess the model's response variables.
• Utilized Equal Units (EU) analysis as a pre-processing step for Quality of Life data.

Main Results:

• The model's expected response variables satisfy additive conjoint measurement axioms for a common scale.
• An EU analysis on Quality of Life data yielded a simple structure model with three dimensions in equal units.
• The derived scores enable addressing research questions previously unanswerable with unequal units.

Conclusions:

• When the Equal Units (EU) model fits data, it provides enhanced validity for psychological measurements.
• The validated EU scores allow for more meaningful within-person comparisons.
• Further research is needed to explore limitations and applications of this measurement approach.