Even the same expressionless face was rated slightly more pleasant after people smelled jasmine than after clean air alone. After an odor resembling rotten cabbage, by contrast, ratings were lower.
A 2015 study examined not only this shift in impressions but also brain waves recorded while people viewed the faces. Its distinctive feature was that the odor and face were not presented together: the researchers waited one second after the odor ended before showing a face. They asked whether an immediately preceding olfactory experience could carry over into evaluations of the next face, and at what stage of information processing any differences would appear.
Showing a face one second after the odor ends
The researchers screened 23 people beforehand, and 20 who met the criterion on an odor-identification task took part in the experiment. Their average age was about 25. The visual stimuli were 36 photographs of neutral faces, comprising 18 men and 18 women.
There were three odor conditions: a pleasant jasmine fragrance, unpleasant methylmercaptan, and clean air as a control. A constant stream of air was delivered through tubes near the nose, with the odor introduced for just three seconds. This arrangement helped avoid confusing the sensation of a sudden change in airflow with an odor effect.
One second after odor presentation ended, a face photograph appeared for 300 milliseconds. After a further 1.7-second wait, participants rated the face’s pleasantness from 0 to 100, then rated the intensity of that trial’s odor. A 129-channel EEG system, along with measures including respiration, recorded responses while they viewed the faces.
The task consisted of three blocks of 36 trials, for a total of 108. Each face appeared once in each block and, across the experiment, was paired once with each of the three odor conditions. This design allowed the same face to be compared across odor conditions without the comparison depending on differences between faces.
Face pleasantness: jasmine first, then air, then the unpleasant odor
Mean face-pleasantness ratings were 55.26 in the jasmine condition, 53.19 with air, and 50.19 with methylmercaptan. All three comparisons were statistically significant: jasmine was rated higher than both air and the unpleasant odor, and the unpleasant odor lower than air.
On the 100-point scale, the difference from air was about two points for jasmine and three points for the unpleasant odor. Rather than dramatically reversing an impression, the preceding odor produced a small shift in the perceived pleasantness of the same face. This was not a test of facial trustworthiness, personality, or attractiveness as a romantic partner; it measured how pleasant or unpleasant a face seemed.
The two fragrances had been matched approximately for intensity in a pilot test, but measurements before and after the main experiment showed that methylmercaptan was perceived as stronger. During the face-rating task itself, however, the difference in intensity ratings between the fragrances was not significant. The conditions should therefore be understood as retaining some difference in intensity as well as in pleasantness.
Brain-wave differences emerged about 0.6 seconds after the face appeared
The study’s other main focus was event-related potentials: brain-wave responses averaged in relation to face onset. The researchers tracked the time course after a face appeared, exploring when and where on the scalp differences between odor conditions emerged.
At approximately 621–633 milliseconds after face onset, electrodes near the right parietal region showed a more negative potential in the jasmine condition than in the air and unpleasant-odor conditions. The difference between air and the unpleasant odor was not significant. The authors discussed the possibility that a pleasant odor affected relatively late processing associated with evaluation and emotion.
At around 926 milliseconds, further differences appeared at lateral frontotemporal sites on both sides. On the left, jasmine produced a more positive potential than the unpleasant odor; on the right, the unpleasant odor and air produced more positive potentials than jasmine. However, neither the left-sided jasmine-versus-air comparison nor the right-sided unpleasant-odor-versus-air comparison was significant. It is more appropriate to retain these condition-specific comparisons than to reduce the finding to a simple division of pleasantness on the left and unpleasantness on the right.
Respiratory movements also differed between conditions, but an analysis taking them into account found no significant evidence that breathing explained the EEG differences. Nor was there a significant correlation between face-pleasantness ratings and EEG amplitudes. The study therefore did not establish that people with larger brain-wave changes also experienced larger shifts in their impressions of faces. One negative correlation remained between the unpleasant odor’s own pleasantness, measured before and after the task, and the left-sided potential at around 926 milliseconds, but the authors interpreted this association cautiously as well.
An exploratory EEG study of impressions carrying over
Because face ratings changed even with a one-second gap between odor and face, the authors interpreted the findings as positive or negative responses to an odor carrying over into the subsequent evaluation of a face. The design makes it harder to explain the result as participants simply reporting the pleasantness of an odor they were smelling at the same time.
The EEG findings are exploratory evidence that can guide further research. The researchers first searched many time points and electrodes using a threshold uncorrected for multiple comparisons, then selected locations showing differences and reanalyzed the same data. The later analysis was not an independent replication, making it important to test the specified locations and timings in another experiment.
The original paper requires care when interpreting how effects changed with repeated exposure. In the block-by-block results, one passage describes an effect as significant while giving p = 0.14. The discussion’s suggestion that the unpleasant odor’s influence weakened also does not align with the decrease in face ratings reported in the results. Although the overall comparisons among the three conditions are presented, the direction of habituation-related changes should not be regarded as established by this report.
Summary
The same faces were rated more pleasant after jasmine than after air, and less pleasant after the unpleasant odor. A key finding is that a shift in impressions consistent with the preceding olfactory experience remained even when faces appeared one second after the odor ended.
Differences between odor conditions were also found in brain waves around 0.6–0.95 seconds after face onset, suggesting a possible role in later evaluative processing. However, face ratings did not correspond directly to EEG amplitudes, and the analysis was exploratory. The study captures, through both subjective ratings and EEG, how a brief olfactory experience can carry over into a subsequent visual judgment.
Research context and disclosures: The study received an ESRC CASE studentship and additional support from Unilever R&D. Two coauthors were Unilever employees at the time. The paper states that the odors used were not present in the company’s commercial products. The other authors declared no commercial or financial relationships that could constitute a potential conflict of interest.
Reference
Cook, S., Fallon, N., Wright, H., Thomas, A., Giesbrecht, T., Field, M., & Stancak, A. (2015). Pleasant and Unpleasant Odors Influence Hedonic Evaluations of Human Faces: An Event-Related Potential Study. Frontiers in Human Neuroscience, 9, 661. Original paper / Open full text
