Vanilla feels comforting, while mint feels fresh and invigorating. Do these associations with scents appear not only when people take time to answer in words, but also in tasks that require a quick key press?
In a 2019 study, people classified scents and feeling-related words faster when “vanilla and relaxation” and “mint and energy” shared the same response keys than when the pairings were reversed. This difference appeared both with scent names alone and when people smelled actual odors without being told their names.
- Using reaction times to investigate associations before people put them into words
- Starting with just the names “vanilla” and “mint”
- Responses were faster even when the odors were unnamed
- Testing two blended fragrances as well as single ingredients
- Associations with relaxation are not the same as actually becoming calmer
- Summary
Using reaction times to investigate associations before people put them into words
Feelings about scents are usually investigated with questionnaires. However, answers also depend on people’s ability to describe their experiences and their ideas about how they should respond. As a complement to self-report, the researchers asked whether they could measure automatic associations between scents and relaxation or energy.
They adapted a task called the Implicit Association Test, or IAT, for odors. Participants classified words appearing on a screen or odors delivered to their noses as quickly and accurately as possible, using one of two keys.
For example, one task assigned “vanilla” and words meaning “calm” to the left key, and “mint” and words meaning “energetic” to the right. Another paired vanilla with energy and mint with relaxation.
If vanilla is associated with relaxation, the first assignment should be easier to process, while the second should take a little longer. The researchers used this difference in reaction time between pairings as evidence of an association.
Starting with just the names “vanilla” and “mint”
Experiment 1 used no scents, presenting only the words “vanilla” and “mint” on a screen. The relaxation category contained French words meaning relaxed, calm, or reassured, while the energy category included words meaning energetic or refreshed.
Of 32 participants, 31 were included in the IAT analysis. Mean reaction time was 776 milliseconds for the expected pairings and 910 milliseconds for the reversed pairings, a significant advantage for the former. These were overall task averages combining trials with scent names and trials with feeling-related words.
In a questionnaire after the task, the name vanilla was also associated more strongly with relaxation than mint, while mint was associated more strongly with energy than vanilla. At this stage, however, the findings concerned relationships between words. The next experiment tested whether the same associations would appear when people actually smelled the odors.
Responses were faster even when the odors were unnamed
Experiment 2 used vanillin, which smells like vanilla, and menthol, which smells minty. A device delivered the odors to participants’ noses, and they learned them as “odor 1” and “odor 2,” without being told their names. Tasks distinguishing the two odors were combined with tasks classifying feeling-related words.
Of 31 participants, 30 were included in the IAT analysis. Mean reaction time was 1,072 milliseconds when vanillin and relaxation shared a key and menthol and energy shared the other, compared with 1,359 milliseconds for the reversed pairings. Again, the former was significantly faster. These averages also included word trials, not just odor trials.
In self-ratings, vanillin was judged more relaxing and menthol more energizing. Menthol was also rated stronger and more familiar, so the stimuli were not matched in every respect.
Even without a name, people may recognize a familiar scent and think, “This is vanilla.” The researchers therefore moved on to scents that were harder to identify by name.
Testing two blended fragrances as well as single ingredients
Experiment 3 used two perfume compositions. One was a citrus-aromatic fragrance containing notes such as bergamot and lemon; the other was an oriental-floral fragrance combining flowers with notes such as vanilla and musk. They were selected from prior evaluation data as, respectively, energizing and relaxing fragrances.
Thirty-five people participated, and the reported IAT analysis included 34. These fragrances were also presented without names. Mean reaction time was 1,244 milliseconds for the expected assignments and 1,364 milliseconds for the reversed assignments, a significant difference.
Self-ratings also distinguished the two fragrances in relaxation and energy. In contrast, differences in intensity, familiarity, and pleasantness were not significant. Thus, task-based associations appeared not only for simple fragrance ingredients, but also for more complex compositions that were less familiar than vanillin and menthol.
In all three experiments, overall accuracy did not differ significantly between the pairings. An analysis comparing standardized reaction-time differences across experiments also found no significant difference. This does not, however, prove that all three effects were exactly the same size.
Associations with relaxation are not the same as actually becoming calmer
This study measured how quickly people could combine scents and feeling-related words. It did not test whether mood improved, concentration increased, or the body became calmer after smelling a fragrance.
The authors discuss the possibility that repeated pairings of scents and situations in daily life lead to learned associations. The effect of names alone is consistent with this explanation. However, the experiments did not teach participants new associations, so learning cannot be established as the sole cause.
Participants were French or Swiss nationals recruited from different departments at Firmenich. The researchers did not check whether they recognized the fragrances. Further work is needed in groups with different cultures and experiences, and with scents that are less clearly relaxing or energizing.
Odors are also difficult to present in rapid succession, so this task was not identical to a conventional, predominantly visual IAT. Rather than a tool that directly reads a person’s true feelings or unconscious mind, it is better viewed as another way to investigate associations that questionnaires alone may miss.
The study included authors affiliated with Firmenich and reported support from the company and Swiss research funding. The authors declared no potential conflicts of interest concerning the writing and publication of the article.
Summary
Across three experiments using the names vanilla and mint, actual vanillin and menthol, and two blended fragrances, responses were faster when expected relaxation and energy pairings shared a key than when those pairings were reversed.
The findings support automatic associations between scents and feelings that appear during rapid classification. They also suggest that an odor-adapted IAT could complement self-report in research. Distinguishing “associated with relaxation” from “actually calming the mind and body” allows the effects of scent to be investigated more precisely.
References
Lemercier-Talbot A, et al. Measuring Automatic Associations between Relaxing/Energizing Feelings and Odors. Food Quality and Preference. 2019;77:21–31. doi:10.1016/j.foodqual.2019.04.010
