On the validity of measuring brand images by rating concepts and free associations

Abstract

This article provides a validity assessment of measuring brand image by free association using newly developed Web-based tools. Nomological validity is assessed by testing causal effects as predicted by a causal model for brand commitment when brand image is measured by rating free associations. Three sets of data were collected to assess validity: (1) traditional questionnaire to measure the construct indirectly with multiple indicators, (2) traditional questionnaire with direct construct assessment and (3) a virtual ideation session with direct construct assessment. Estimable causal effects were identified by Pearl's ‘back-door’ criterion; the nomological network was found to be valid: select causal predictions of the brand commitment model held in all three measurement conditions. Measuring brand image by rating concepts and free associations show promise in better understanding consumers’ brand images and brand equity, as well as in manager's efforts at brand positioning and repositioning. Brand image as traditionally measured by free association is labor intensive, time consuming and expensive. New Internet tools such as dialogr.com show promise of alleviating these impediments and of delivering a deeper understanding of what consumers think and how they feel about brands.