A sprinkle of emotions vs a pinch of crossmodality - DATASET

Published: 07-02-2020| Version 1 | DOI: 10.17632/rrm7sbfpcy.1
Felipe Reinoso-Carvalho,
German Molina,
Laura Gunn,
Enrique ter Horst,
Johan Wagemans,
Takuji Narumi,
Yuji Suzuki,
Charles Spence


We report a study designed to determine the most efficient means of pursuing sonic seasoning strategies, while cross-culturally comparing music that was chosen to trigger a specific emotional response, versus music that had been chosen to be crossmodally congruent with specific taste/flavor attributes. The effects triggered by ‘emotional’ music were found to be more prominent than those triggered by ‘crossmodally-corresponding’ music. Specifically, a chocolate was liked more, rated as sweeter, and the purchase intent was higher, when it was tasted with positive, as compared to negative, emotional music. By contrast, the same chocolate was mostly rated as tasting more bitter with the negative, as compared to the positive, emotional music. Those companies wanting to effectively use sonic seasoning techniques as part of their marketing strategies, should therefore principally aim at intelligently classifying music based on the emotions that they can trigger in consumers (at least when thinking more globally). == DATA ANALYSIS OVERVIEW: A multivariate multinomial ordered probit model was implemented to measure the differences between ordinal scales, with a dummy intercept vector to account for the within-participant measurement changes in chocolate perception across measured ordinal items, while also accounting for covariate differences (gender, age, order effects). In order to measure the relative willingness to pay, the ratio of quantitative measurements was log-transformed to represent a percent difference between tasting (with the baseline tasting in the denominator), and regressed against the same set of covariates defined above assuming a Gaussian relationship. The latter statistical analysis was performed using the core syntax of R. Hirk, R., Hornik, K., & Vana, L. (2019). Multivariate ordinal regression models: An analysis of corporate credit ratings. Statistical Methods & Applications, 28(3), 507-539. Hirk, R., Hornik, K., Vana, L., & Genz, A. (2019). Multivariate Ordinal Regression Models [computer software]. GPL-3 licensing. Available from https://cran.r-project.org/web/packages/mvord/mvord.pdf