To explore whether gelati and music pairs were significantly
related to the emotional states of the panellists (the dependent
variable), and pleasantness ratings of gelati, a full factorial ANOVA
with gelati (3 levels) and music (4 levels) as within groups factors
was performed, with post hoc Tukey's analysis applied if significance
was observed. Superscripts were added when significant
differences either between music samples (disliked, neutral, and
liked) within the same gelato flavour (DC for example) or between
gelato flavour (DC, BC, and MC) within the same music sample
(disliked music for example) were found similar to Kantono et al.
(2015).