Comparative Analysis of Factor-Based and Composite-Based Structural Equation Models: Factors Affecting Word-of-Mouth of Tourists in Khung Bang Kachao

Abstract

Structural Equation Models (SEMs) are commonly categorized into two main types: factor-based SEM and composite-based SEM. Each type is suitable for analyzing different forms of hypothetical constructs. Factor-based SEM is well-suited for analyzing factors, while composite-based SEM is tailored for analyzing composites. However, the majority of past research has favored composite-based SEM, particularly Partial Least Squares (PLS), for analyzing factors. Such practices can introduce biases into the analysis.

This article provides an illustrative example from the tourism and hotel management domain by analyzing hypothetical constructs in two scenarios: one where the constructs are treated as factors and another where they are treated as composites. The study includes six constructs: service quality, atmosphere, perceived value, satisfaction, revisit intentions, and word-of-mouth. In the first scenario, the research objective is theory testing, while in the second scenario, the research aims to assess the model’s predictive capabilities when applied to datasets beyond those used for the analysis.

The constructs of service quality, atmosphere, and perceived value are assumed to influence satisfaction, while satisfaction and service quality are hypothesized to impact revisit intentions. Perceived value, satisfaction, and revisit intentions are further assumed to trigger word-of-mouth.