An Investigation of Patterns of Self-Brand Personality Alignment

Karampela, Maria and Tregear, Angela (2015) An Investigation of Patterns of Self-Brand Personality Alignment. PhD thesis, University of Edinburgh.

Full text not available in this repository.Request a copy


In their continuous search for improved explanations of why consumers engage with certain brands more than with others, academics and managers have placed significant importance on self-brand congruence theory, which suggests that consumers are drawn to brands with characteristics that align with their own characteristics. Although this theory has been extensively investigated and supported by previous research, it has hitherto been assumed that the alignment of characteristics exhibits a similarity configuration, that is, consumers are drawn to brands with traits that mirror their own traits. By adopting a relational view of consumer-brand interactions and drawing from the literature on interpersonal relationships, specifically from theories of interpersonal attraction, the thesis explores the possibility that besides similarity, self-brand personality alignment may also exhibit a complementarity configuration, whereby consumers are drawn to brands with traits that complement their own. Using a two-phase empirical study of mixed methods [in-depth interviews and online survey (n=206)], the thesis first explores the structure of consumers’ perceptions of their own personalities and those of their favourite brands using Exploratory Factor Analysis. Results reveal that although brand personality has the same five factor structure as human personality, the composition of dimensions is distinct. The patterns of alignment between the two sets of traits are then examined through Canonical Correlation Analysis, which reveals the existence of both similarity and complementarity configurations in self-brand personality alignment. A new method for measuring the magnitude of self-brand personality alignment is then devised, which captures both configurations. The predictive power of this new measure is then compared against existing, similarity-based measures for a range of desirable brand behaviours, using Discriminant Analysis and Linear Regression. Results indicate the new measure performs well, especially for emotionally-related brand behaviours.


Karampela, Maria ORCID logoORCID: and Tregear, Angela;