Structural Equation Modeling for the Interplay Among Consumer Engagements with Multiple Engagement Objects in Consumer's Fashion

Consumer engagement (CE) is attracting much attention from academics and practitioners. Early research on CE focused on online contexts such as an online brand community. Recent research attention of CE is not only in online but also in different offline settings. The research is mainly conceptual and is concerned about CE with a single engagement object. However, there is little understanding about CEs with multiple engagement objects. We explore how consumers engage with multiple engagement objects in a fashion consumption context and empirically examine the interplay among CEs with them, simultaneously. We define CE by consumer's cognitive, emotional and behavioral investments into fashion-related interactions. Both qualitative and quantitative research are conducted from this CE perspective. In the qualitative research, we draw on semi-structured interviews and extract three focal objects. This research results are used to develop conceptual models of the interplay among CEs with these objects. The quantitative research collects data for measuring CEs with Likert type scale. Structural equation modeling (SEM) measures how well the conceptual models fit the data. Moreover, we propose a new approach NPCA-SEM that includes nonlinear principal component analysis (NPCA) in SEM. NPCA finds optimally scaled data quantifying Likert type scale data and computes principal component scores of the optimally scaled data. SEM treats the scores as observed data instead of Likert scale data. In the analysis of our research data, we compare the goodness of fit of conceptual models from NPCA-SEM with that from ordinal SEM and examine the performance of NPCA-SEM.