Linking value, confirmation and satisfaction to predict behavioural intention : examining alternative models in a service environment

University of Technology, Sydney. UTS Business School.The rapid advancement of the services sector has recently led to significant changes in the global economic structure. Consequently, academic researchers and service practitioners are focusing on better delivery of consumer value and increasing consumers’ purchase intentions. Unfortunately, services marketing and consumer behaviour literature has been restricted to models for replication or model development rather than alternate model comparison. To fill this gap, this study has examined three alternate models in a service environment to advance the knowledge of behavioural intention determinants from the consumer perspective. This study proposes value-based features and an expectancy–confirmation theoretical framework in a model which it compares with two prior classic models. Unlike most prior studies, this study reconceptualizes the perceived value construct from the multi-dimensional perspective for its proposed model. The Best–Worst Scaling (BWS) method is applied to measure this construct filling existing knowledge’s methodological deficiency. Due to the importance of tangible and intangible features in the restaurant environment, the Australian restaurant services sector has been considered an ideal research setting for exploring inherent utilitarian and hedonic value dimensions. For data collection, this study used a web-based survey by an online research organization, and structural equation modelling (SEM) with AMOS software was chosen as the major data analysis tool. The empirical findings confirmed Alternate Model 2 as a better model in predicting consumers’ behavioural intention and found increased acceptability of this model, originally tested in information systems, in services marketing and consumer behaviour literature. An unexpected result concerned the ipsative data problem (a common score for all individuals) of the BWS method resulting in a poorer model fit for the proposed model; however, this study’s methodological contribution involved exploring the BWS method’s hidden data problem. This study also explored the importance of situational impacts on consumer behaviour in model testing. Finally, due to the proposed model’s poorer fit, hierarchical cluster analysis was run on the personal values and perceived value constructs (important constructs of the proposed model): results confirmed that consumers can be segmented based on their personality measures and value preferences. Moreover, multi-group analysis using different clusters explored the significance of developing path-by-path hypotheses in future across different consumers. Despite its limitations, the findings of the study are expected to have substantial implications for academics/researchers and practitioners in service-providing firms: specifically, this study may conceivably produce an agenda for industry-specific improvements in restaurant performance