A Rasch Analysis of Service Performance in a Tourism Context

Advances in tourism research depend on the development and validation of measurement scales using appropriate psychometric techniques. However, scale development practice is dominated by classical test theory, despite its limitations. A Rasch analysis of international travel service performance scales across demographic, customer experience and service provider sub-groups is provided. Scales generally show high reliability, little disordering of the thresholds, and no substantive effects where multidimensionality is present. Some items exhibit limited discriminatory power, local dependence and measurement bias. Measurement invariance violations are prevalent in temporal dimensions where travelers exhibit diverse customer experience profiles. Employee knowledge and safety attributes appear organization-specific. Ethnicity impacts on relational aspects; response behavior regarding personal space is gender specific. Standardized safety-related scales may be inadvisable when comparisons across age and education cohorts are required. Accounting for measurement invariance influences the average location of groups on the latent continuum, and in some cases substantive mean difference results.