What if you kept the same scent near you during sleep that you smelled while learning vocabulary? Could that cue make it easier to recall the words later?
In a 2020 study, 54 German students learned English vocabulary in their usual school and home settings. The group that used a rose scent while studying and sleeping made fewer errors on a test than the group that used no scent. The study investigated an effect previously observed in laboratory sleep research under conditions closer to everyday learning.
Using a scent as a cue for a learned memory
Earlier research had reported that presenting an odor while people learned something, and presenting it again during sleep, could help consolidate the associated memories. This approach is called targeted memory reactivation.
The idea is not that the scent itself supplies the answers, but that it acts as a cue associated with material already learned. The approach investigates whether the process of organizing memories during sleep can be influenced from outside.
The study addressed three questions: whether the effect occurs in real-life school and home settings; whether it occurs when a scent is present throughout the night rather than timed to sleep stages measured with brain-wave recordings; and whether adding the same scent during a later test further improves performance.
Varying when scent was used in two classes
The participants were 54 German students aged 11–12, in two classes of 32 and 22. All had already received one year of English lessons. Within each class, half were randomly assigned to the scent group and half to the control group.
The vocabulary was first introduced in school lessons without scent. Students then studied at home and took a test at school one week later. The tests included German-to-English and English-to-German translation, sentence construction, and sentence completion. Spelling mistakes, translation mistakes, and failures to recall answers were counted as errors.
The study used the following conditions in sequence.
- No scent: Both groups studied and took the test without scent.
- Learning + test: Only the scent group used the rose scent during home study and the school test.
- Learning + sleep: Scent was used during home study and sleep on all seven nights of that week, but not during the test.
- Learning + sleep + test: Scent was used in all three settings.
The control group used no scent during any period. For school-related reasons, the final condition involving all three settings was carried out only in the class of 32 students.
The researchers used commercially available rose-scented incense sticks, but they were not lit. The sticks gave off a scent without burning and were placed on the study desk, near the bed, or on the test desk. During sleep they remained in place throughout the night, and sleep stages were not measured.
Fewer errors when scent accompanied learning and sleep
Combining the data from all 54 students in the two classes, the mean error rate in the learning + sleep condition was 25% in the scent group and 36% in the control group. The group difference remained significant after correction for multiple comparisons.
In the learning + test condition, by contrast, the rates were 32% and 34%, with no significant difference. The no-scent period also showed no clear group difference, with rates of 33% and 32%.
Together, these results suggest that under the tested conditions, a benefit appeared when the same cue used during learning was also present during sleep, rather than simply matching scent at learning and testing. They indicate that memory cueing during sleep is worth studying in everyday learning.
However, the different conditions involved different vocabulary, and everyone experienced the conditions in the same fixed order. Within a condition, both groups learned the same words, so comparisons between groups are possible. But comparing 25% in one condition with 32% in another does not, by itself, establish which way of using scent is better.
Does adding scent during the test help even more?
In the learning + sleep + test condition, conducted in the single class of 32 students, error rates were 18% in the scent group and 35% in the control group. The group difference within this condition was also significant.
That alone does not demonstrate an additional benefit from adding scent at testing. In the same class, the interaction testing whether the difference from the control group increased when moving from learning + sleep to all three settings was not significant.
The paper identifies the possible benefit of additional cueing as a question for future work. The 25% versus 36% figures for learning + sleep are averages from two classes, whereas the 18% versus 35% figures for all three settings come from one class. Calculating an improvement by directly comparing those pairs therefore also requires caution.
What the result with all-night presentation means
Sleep studies have often recorded brain waves and delivered odors during particular stages, such as deep non-REM sleep. The present method was simpler: a scented object remained in place throughout the night. Observing a performance difference even with that approach was an important finding.
However, the study cannot identify when during the night the cue was effective. Nor did it directly compare all-night presentation with presentation limited to a particular sleep stage, so it does not establish equivalent effects or an optimal method.
The proposed mechanism is that an odor associated with learning helps reactivate memories during sleep. Because this study did not record brain activity, it used differences in performance to examine whether such a mechanism might also operate in everyday life.
Questions that remain in a real-life study
The researchers did not control how many hours students studied at home, how far away they placed the scent, or their actual sleep duration and quality. Scent use was also assessed through self-report.
The control group did not receive an identical unscented object. Expectations from knowing that one was using a scent, or changes in study behavior, may therefore have contributed to the results. The two classes also differed in teachers and details of their tests.
These conditions illustrate the challenges of research outside the laboratory as well as opportunities for improvement. Studies with more participants, balanced condition orders and study time, and attention to expectations would clarify how far odor cues can be put to use.
Summary
Students who used a rose scent while learning and during sleep on the following seven nights had a lower vocabulary-test error rate than students who used no scent. Importantly, the study showed that memory cueing could be investigated under everyday conditions, with scent present throughout the night and without measuring sleep stages.
The condition that also included scent during testing outperformed the control group, but any extra benefit of that additional presentation remains uncertain. The study points to a possible aid for using material already learned. Learning the material in the first place remains the foundation for exploring the relationship between scent and sleep.
Publication costs were supported by the German Research Foundation and the University of Freiburg. The authors declared no competing interests.
References
Neumann F, Oberhauser V, Kornmeier J. (2020). How odor cues help to optimize learning during sleep in a real life-setting. Scientific Reports, 10, 1227. Original article / Open-access full text
