Convergent validity in tourism research: an empirical analysis

Correlations between the measures were generally low, although convergence at the aggregate level may have been higher than at the individual level, and caution should be exercised by tourism researchers in relying on single techniques of assessing attribute importance.

Abstract

In order to examine the convergent validity of several techniques, four measures were obtained of the influence of various attributes of computerized reservation systems in purchase decision making by travel agents. Measures were gathered at the individual respondent level. The four measures were: (1) elicitation; (2) selective ranking; (3) direct rating; and (4) conjoint measurement. In order that the convergent validity of this conjoint model with another technique could be assessed, data for a self-explicated multi-attribute model were also gathered. Correlations between the measures were generally low, although convergence at the aggregate level may have been higher than at the individual level. Caution should be exercised by tourism researchers in relying on single techniques of assessing attribute importance.