BRIEF REPORT: ADDITIONAL DATA ON DISTORTION DUE TO AGGREGATION IN NONMETRIC MULTIDIMENSIONAL SCALING

Area of Science:

• Psychology
• Data Analysis
• Multidimensional Scaling

Background:

• Previous research indicated aggregation had minimal impact on nonmetric multidimensional scaling (NMDS) solutions when stimuli were adjectives.
• Recent studies using nonadjectival stimuli suggest aggregation can introduce significant distortions in NMDS configurations.

Purpose of the Study:

• To investigate the impact of data aggregation on nonmetric multidimensional scaling solutions.
• To explore potential differences in aggregation effects between adjectival and nonadjectival stimuli.
• To underscore the importance of aggregation effects for the validity of multidimensional scaling analyses.

Main Methods:

• Analysis of similarity data from previous research involving adjectival stimuli.
• Replication of aggregation effects using nonadjectival stimuli.
• Comparison of aggregated and individual similarity data configurations.

Main Results:

• Aggregate solutions closely matched individual solutions for adjectival stimuli.
• Substantial distortions were observed in aggregate solutions for nonadjectival stimuli.
• Preliminary hypotheses were formulated regarding differences between adjective and object stimuli.

Conclusions:

• Aggregation effects in nonmetric multidimensional scaling are stimulus-dependent.
• Findings suggest caution when aggregating data for nonadjectival stimuli in NMDS.
• The study highlights the critical need to consider aggregation methods in multidimensional scaling.