Quantitative results

Matches
Participants were asked to rate the match of the wines with the soundscapes in each section on the 7-point scale, with 1 as ‘not well matched’ and 7 as ‘well matched’. For the sparkling wine with the soundscape for S1, the mean rating was 5.41, with a standard deviation of 1.35. The Pinot Gris had a mean of 4.64 with a standard deviation of 1.94. Finally, for the red blend the mean rating was 4.43 with a standard deviation of 2.06.

An analysis of variance was performed to determine whether these ratings indicated a significant agreement that the wine and soundscape matched. The null hypothesis was that observers would think the pairings were neither well matched nor not well matched and would thus, on average, give a rating of 4. This value was therefore subtracted from each rating, so that a positive intercept would indicate evidence that the pairings were judged to be significantly well matched. The type of wine (sparkling wine, Pinot Gris or red blend) was used as a within-subject factor and the location (London or Sydney) as a between-subject factor. A significant intercept (F(1,50)=9.9, p=0.003) indicated that the pairings were judged to be well matched. There was no significant effect of city (F(1,50)=1.1, p=0.290) indicating group agreement between the two studies. There was a significant effect of wine (F(2,100)=4.9, p=0.009) indicating that some pairings were judged as better matched than others. Post-hoc pairwise comparisons showed that the match for the sparkling wine was judged as better than that for either the Pinot Gris (p=0.020) or the red blend (p=0.004). There was no difference in the degree of match between the Pinot Gris and red blend (p=0.697).

Change
Participants were asked to rate the level of change to the taste/texture of the wine with the section on a 7-point scale on which 1 was ‘noticed no change’ and 7 was ‘noticed much change’. For the sparkling, the mean change noted was 4.05, with a standard deviation of 1.87. The Pinot Gris had a mean of 3.57 and a standard deviation of 1.93. The red blend had a mean of 3.59 and a standard deviation of 1.84.

Analysis of variance was also used to assess the statistical significance of the apparent change in perception. In this case, no change would be indicated by a rating of 1, so this value was subtracted from all of the ratings. A significant intercept (F(1,50)=253.2, p<0.001) indicated a significant change in perception overall. There was no significant effect of city (F(1,50)=1.0, p=0.312) indicating group agreement between the two studies. There was a significant effect of wine (F(2,100)=7.7, p=0.001) indicating that some of the changes in perception were judged as stronger than others. Again, post-hoc pairwise comparisons showed that the change in perception was greater for the sparkling wine than for either the Pinot Gris (p<0.001) or the red blend (p=0.042). There was no difference in the change in perception between the Pinot Gris and red blend (p=0.088).

Expertise
No significant effect was found for city, despite differences in the balance of expertise in music and wine between the two groups.