Comparison of explicit vs. implicit measurements in predicting food purchases

Abstract In this study, we aimed to investigate the relation between consumer purchases of three branded blueberry flavored quarks and respective responses of the same consumers to these products using 1) traditional explicit consumer surveys measuring verbalized impressions, 2) novel explicit pictorial emoji scores and 3) implicit behavioral responses produced during an approach-avoidance task (AAT). Explicit measures (n=134) were collected before product tasting (expectation condition) during an online survey, and after product tasting (perception condition) during a Central Location Test (CLT). Implicit measures were collected with a subset of 56 randomly selected subjects during the CLT. These included electroencephalographic (EEG) measures, joystick response speed and pupil size responses. During one month following the CLT, respondents registered their purchases via an online diary. Bivariate correlations indicated that explicit scores correlate better with product purchase amounts in the perception condition than in the expectation condition. Furthermore, verbalized ratings correlated better with product purchase amounts than pictorial emoji scores. Of the implicit responses, EEG responses produced the strongest correlations with purchase behavior, similar to those observed for verbalized explicit ratings in the expectation condition. Multiple linear regression modelling indicated that the best-fitting model consisted of an emoji score, purchase intention score, pleasantness score, brand relationship score, and implicit joystick response speed. Overall, purchase behavior was associated stronger with explicit responses than with implicit responses. Yet, the prominent role of implicit joystick response speed in the multivariate regression model suggests its unique contribution to the understanding of purchase behavior.