Isolation Booths, Immersive Technology, and Wine Bars: How Does Environment Influence Consumer Wine Preference?

Often in research settings, you want to control for as many variables as possible, and rightfully so, for statistical purposes.  For example, in experiments involving tasting of wine, participants might be placed in isolation booths with minimal external stimuli present (no sound, black glasses, no packaging present, etc.).  While this might yield results that are statistically sound, are the results really what would happen in real-world scenarios? Correct me if I’m wrong, but people don’t usually sit around drinking wine in silent isolation booths.  There is often music playing, social interactions, and all sorts of external stimuli, many of which have been shown to influence ones liking of any given wine.

On the other hand, performing a study in the middle of a restaurant or noisy wine bar could be very complicated, as now one has introduced countless other variables that are not under experimental control, and thus much more difficult to factor in when performing statistical analysis.

So what’s the best approach here?  Wine tasting in isolation booths where it may or may not reflect how a person would behave in a real-world scenario? Or wine tasting in a real-world setting where there are many uncontrolled variables that can muck up the statistical analysis?

Photo courtesy pixabay.com

In recent years, a third option has become available to researchers of sensory-type experiments, and that is using immersive technologies like virtual reality (etc.). Perhaps by using immersive technologies researchers can control certain variables while still providing a more “real-world”-type experience than the simple isolation booth would.

A new study, available online and to be published in the September 2019 issue of Food Quality and Preference, aimed to examine the use of immersive technology in a wine setting, specifically by comparing a virtual wine bar with an actual wine bar setting as well as an isolation booth setting. This comparison would hopefully determine if wine preference/liking was similar (or not) to the two wine bar set-ups and therefore more reflective upon real-world situations instead of the isolation booth experimental design.

Brief Methods

Note: I’m leaving out a lot of specifics for space consideration.  If you have specific questions, just ask and I’ll see if I can find the answer in the original research.

To perform this experiment, three different tasting sessions occurred:  1) isolated tasting in a controlled laboratory environment, 2) an immersive environment, and 3) a real wine bar environment.

For the controlled laboratory environment, the traditional isolation booth method was utilized and only one participant was present at a time.

For the immersive environment, participants again tested alone, however this time, they were tasting wines in a room filled with large TVs running footage of an actual wine bar.

For the real wine bar, participants tasted the wines at Wine On High in Columbus, OH, where it was closed to the public for the duration of the experiment. This time, up to 10 participants were present at the same time, to represent what one would likely come across in a social setting. Participants were encouraged to socialize, but not about the wines that they were tasting.

immersive technologies
Fig 1 from Hannum et al 2019

Four different wines were used in this study, all cabernet sauvignon representing a wide range of price points: 1) 2016 Black Box; 2) 2016 1895; 3) 2014 St. Michelle; and 4) 2014 Freakshow.

Participants were asked to rate their overall liking of each wine using a 9-point hedonic scale.  Participants were also asked to specify their level of agreement on a 5-point scale with two statements: “I would order this wine at a wine bar or restaurant” and “I would purchase this wine to drink at home”.  Finally, participants were asked to guess the price of each bottle by selecting one of five price range options.

Participants were eligible for this study if they consumed red wine at least twice per week, consumed wine either at home or in a restaurant within the past month, had never been to Wine On High, and hadn’t participated in a sensory study in the past month.

A total of 62 people (18 men, 44 women) participated in the study, their ages between 23 and 59 years old.

Selected Results

  • 75%+ of participants said they drank red wine at home much more often than in a wine bar or restaurant setting.
  • 80% of participants said they had at least some knowledge of wine, 7% knew a lot of information, and 10% knew very little.
  • Tasting environment (isolation booth vs immersive environment vs real wine bar environment) only had marginal effects on wine liking, which were NOT statistically significant. 
    • Overall ratings trended higher in the real wine bar setting compared with the immersive setting and the isolation booth, but again, this trend was not statistically significant.
  • St. Michelle and Freakshow wines were liked significantly more than the 1895.
  • Wines were rated similarly across all three environments.
  • While average wine ratings were statistically similar between environments, on the individual participant level the results were a different story.  Specifically, there was a lot more inconsistency in wine rating across the three environments on the individual participant level.
    • St. Michelle wine ratings tended to be the most stable between the environments for individual participants, while Black Box ratings were the most unstable.  This result, however, was NOT statistically significant.
  • For the question of whether or not they’d order a wine at a wine bar, significantly more participants agreed when they were in the real wine bar setting compared with the isolation booth setting.
    • Participants were more likely to order the St. Michelle and Freakshow than the 1895.
    • The wines chosen to purchase for home consumption were similar between the three environments.
  • For guessing wine prices, participants guessed wines were significantly more expensive when they were in the real wine bar setting compared with the isolation booth and immersive setting.
  • Overall, participants were able to correctly discriminate the more expensive wines and the less expensive wines used in the study.

Conclusions

Overall, these results seem to suggest that the testing environment does NOT influence wine liking or preference, at least statistically.  While the trend was that wines were rated higher in a real-world setting (what the researchers dubbed a ‘halo effect’), the results were not statistically significant and thus must be assumed to be statistically similar/the same between the three environments. Of course, one problem here is that the sample size might have been too low to detect statistically significance in the results.  Had there been many more participants, it’s possible this trend could have turned significant. Does being in a real-world setting like a wine bar or restaurant positively influence one’s liking of a wine?  According to these results, technically, no, but it’s trending that way and a larger sample size might give a better answer.

This results also brings up the issue of individual inconsistency.  Was there a trend toward higher wine liking scores as a result of the real-world setting?  Or was it just a simple matter of individual scoring inconsistency from one tasting session to another.  Is one person going to score a wine the exact same way every time?  Probably not.  But how much variance is there? And is that why we see a slight difference in liking scores between environments though they were not statistically significant?  Again, a larger sample size MIGHT help here.

Another potential limitation of the study was that even though the real=world setting occurred at a real wine bar, was it “real enough”?  The bar was closed to the public, and a despite being encouraged to socialize with one another, participants were still instructed to perform very specific tasting tasks.  This could potentially skew the results.

Photo courtesy pixabay.com

No matter what, there is always going to be some disconnect between the experimental set-up and a real-world scenario.  There are MANY external influences on sensory preferences and liking, which a strictly controlled experimental design can never truly replicate in order to achieve statistically significance.  The best researchers can do is get a “baseline” expectation/results, and to make inferences as to how consumer behavior MIGHT chance when presented in a real-world setting.

The use of immersive technologies to help bridge this gap in sensory experimental design may be what’s needed, however, much more research needs to be done (specifically with larger sample sizes and more representative of the larger consumer group as a whole).

Source:

Hannum, M., Forzley, S., Popper, R., and Simons, C.T. 2019. Does environment matter? Assessments of wine in traditional booths compared to an immersive and actual wine bar. Food Quality and Preference 76: 100-108.