Entropy-Based Uncertainty-Aware Exploratory Factor Analysis for Ordinal Data: Application to Tramway Cultural Tourism...

Area of Science:

• Social Sciences
• Data Analysis
• Tourism Research

Background:

• Likert-scale surveys are common in tourism and transport research.
• Conventional analysis methods impose artificial precision on ordinal data, ignoring ambiguity.
• Subjective judgments in experiential contexts often involve transitional evaluations, highlighting the need for uncertainty-aware methods.

Purpose of the Study:

• To develop a parameterized fuzzy-entropy exploratory factor analysis (FE-EFA) framework.
• To enable uncertainty-aware analysis of ordinal perception data from Likert-scale surveys.
• To address limitations of conventional point-valued encoding in capturing subjective judgments.

Main Methods:

• Developed a parameterized fuzzy-entropy exploratory factor analysis (FE-EFA) framework.
• Transformed ordinal responses into parameterized fuzzy membership distributions.
• Incorporated Shannon entropy and Jensen-Shannon divergence for distributional analysis.
• Applied the framework to survey data from Chengdu Tramway Line 2 (N = 1242; 32 indicators).

Main Results:

• FE-EFA identified an additional factor compared to conventional EFA.
• FE-EFA demonstrated fewer cross-loadings and a more differentiated loading pattern.
• Information-theoretic measures showed higher entropy (0.8688) and low Jensen-Shannon divergence (0.0133).
• Entropy-adjusted weighting revealed systematic shifts in indicator importance.

Conclusions:

• The FE-EFA framework enhances Likert-scale analysis with an uncertainty-aware layer.
• It preserves structural stability while suggesting a more differentiated latent construct organization.
• The approach offers an exploratory extension for perception-based evaluation and decision support in tourism contexts.