INDIVIDUAL DIFFERENCES IN PERCEIVED TRAIT RELATIONSHIPS AS A FUNCTION OF DIMENSIONAL SALIENCE

Area of Science:

• Psychology
• Multidimensional Scaling
• Cognitive Science

Background:

• The Carroll and Chang INDSCAL model is a key tool for understanding individual differences in perception.
• Its application to trait similarity data requires validation to confirm its utility and robustness.

Purpose of the Study:

• To assess the applicability and validity of the INDSCAL model for analyzing trait similarity data.
• To compare INDSCAL results with other multidimensional scaling techniques and evaluate the impact of trait set size.

Main Methods:

• Two studies utilized the INDSCAL model with trait similarity data.
• Correlations were calculated between INDSCAL solutions and scalar products data.
• Canonical correlations assessed INDSCAL dimensions against Kruskal's non-metric procedure; subject clustering used dimensional salience profiles.

Main Results:

• High correlations (0.81 and 0.71) were found between INDSCAL solutions and scalar products data.
• INDSCAL stimulus dimensions showed high canonical correlations with Kruskal's non-metric procedure results.
• Dimensional structures were comparable whether using 20 traits or 20 traits within a larger set of 60.

Conclusions:

• The INDSCAL model is applicable and robust for analyzing trait similarity data.
• Dimensional structures are stable regardless of the number of embedded traits.
• Subject segmentation based on dimensional salience identified five distinct groups with unique perceptual relationship patterns.