Evaluation of a sensory and cognitive online training tool for odor recognition in professional coffee tasters

The sense of smell plays an essential role in the sensory evaluation of coffee. However, a precise sensory evaluation also requires good memory and the ability to stay focused and concentrated. This study aimed to investigate the effect of an online sensory and cognitive training program on the odor recognition of 36 coffee aromas. A total of 44 coffee professionals participated in a randomized crossover trial. It consisted of an online training program with different sensory and cognitive tasks over two periods of 6 weeks. The participants were divided into three study groups for data analysis purposes: a Control-first group ( n = 16), a matching Intervention-first group ( n = 16), and a second Intervention-first-residual group ( n = 12) to check for repeatability of the test results. On average, the participants improved their odor recognition score by 15% after 6 weeks of training ( p < .01) and maintained a high-performance level after the control period. A positive correlation between completed training and improvement was observed ( p < .001). The participants significantly improved their ability to detect the 36 Le Nez du Café aromas commonly used in the Specialty Coffee Association (SCA) and the Coffee Quality Institute (CQI) exam

using rapid sensory evaluation methods.The recognition of odors is essential for a coffee professional, as these professionals need to be able to recognize and describe aromas in coffee.However, performing a precise sensory evaluation also requires good memory and staying focused and concentrated.Therefore, it is of interest to apply both sensory and cognitive training to improve sensory performance and adherence to the training program due to variability in the tasks.
The recent understanding of brain plasticity, the ability of the brain to develop, has increased research on cognitive training interventions.Research has disproved the belief that the adult brain is incapable of change: several studies show strong evidence that the brain is capable of learning throughout a person's lifespan and, therefore, cognitive functions can be improved though to varying degrees (Lövdén, Wenger, Mårtensson, Lindenberger, & Bäckman, 2013) (Au et al., 2015;Hindin & Zelinski, 2012;Schubert et al., 2014;Schwaighofer, Fischer, & Bühner, 2015).Nevertheless, applying cognitive training to the field of sensory science is novel.Even though cognitive functions such as memory, staying focused, and concentrated are essential skills when performing sensory assessments (Frank, Rybalsky, Brearton, & Mannea, 2011).
Nevertheless, standard sensory training programs are comprehensive and require many hours of on-site training.Therefore, it will be beneficial for semiprofessional panelists to have access to an off-site online training program that will improve their sensory performance.
This study aimed to evaluate the effect of an online combined 2 | MATERIALS AND METHODS

| Participants
A total of 44 coffee professionals (18 females, 26 males, aged  participated in the study.Written informed consent was obtained from all participants involved in the study prior to the first test day. Each participant completed a questionnaire with regard to background information (age, gender, smoking habits, and professional coffee experience, see Table 1).To ensure a high level of completion, the participants paid to participate in the study, which was presented as a sensory performance course with the goal to improve their sensory skills.

| Active control period: Weekly sensory training tasks
The control period was designed as an active control to be perceptually comparable with the intervention training to induce similar expectations and match participants' levels of engagement and interest.
Each week during the control period, the participants received a suggestion for optional training their sense of taste and smell on their own.No cognitive training was included in the control period.An overview of the weekly training task during the control period is presented in Table 3.

| Odor recognition test
The participants took part in three 6-hr test sessions (Test 1-Test 3) onsite before, in between, and after the 6 weeks of either control period or intervention training.The tests included evaluation of both A 2 cm long coffee filter strip (0.5 Â 2 cm) was dipped into the respective aroma bottle, placed into the coded brown bottle with a pincer, and were stored in the fridge (2-5 C) with the screw-lids tight on for maximum 1 week prior to use.The serving order of the odors tested was randomized between tests and participants.

| Satisfaction survey
A survey using the unidimensional 5-point Likert scale was sent out via email after the last test day to understand the participants' attitudes and opinions about the training program.In the questionnaire, the participants were asked to rate their agreement on the scale from strongly agree to strongly disagree, their satisfaction from very dissatisfied to very satisfied, and finally how likely they were to recommend the course to others from very unlikely to very likely.This was all to understand their views and perspectives towards the training program and self-reported improvement.

| RESULTS
The 44 participants followed the training program and completed all three sensory tests.Hereof 16 participants (7 males, 9 females) T A B L E 3 Overview of weekly training tasks during active control period.

| Adherence to the training program
Overall, the results show that the intervention significantly helps the participants to improve their ability to recognize the 36 Le Nez du Café aromas.However, not all participants completed all tasks every day.
The participants were encouraged to complete all tasks but had the freedom to train, as they preferred.As the training program is an expanding binary series, the total number of training combinations is 64, but not all possible combinations were tried.In Table 6, the training combinations that were followed for 1% or more of the days are completed, whereas on 25% of the days, none of the training tasks were completed.On 7% of the days, the breathing exercise was skipped, whereas on 6% of the days, only the olfactory training was completed.Some participants skipped some of the training tasks, some only completed the first half and others chose only to do the breathing and olfactory training (1% of the days).
How often each training task was completed is seen in Table 7; 74% of the days the olfactory training was completed, the different cognitive training tasks were completed 61-64% of the days, while the breathing exercise was the least popular and only completed 58% of the days.
To understand the relation between amount of training and improvement, Pearson's correlation analysis was conducted.In Table 8, it is shown that there is a significant correlation between the number of days where all training tasks were completed and improve-

| Selection of coffee odors
The Le Nez du café aroma set was selected as a test and training tool; however, not all aromas showed to be beneficial for training and improvement.The odor of lemon was well known already at the baseline test, and therefore no improvement was possible.On the other end, some aromas were too difficult to be learned.Even after 12 weeks of training, the aroma medicinal was only correctly named by 36% of the participants, while blackcurrant was only correctly named by 39%.Roasted coffee only reached 32% correct answersan aroma that by its name is supposed to be very familiar to coffee professionals.The low improvement indicates that these aromas do not work as intended when used in an experimental setup with coffee professionals; this is something that SCA and CQI should be aware of before implementing them in their sensory exams.Familiarity is essential for correct odor identification and may explain why some odors were more accessible to learn than others (Knaapila, Virtanen, Yang, Lagström, & Sandell, 2017).For example, the wellknown odors clove, pepper, potato, and dark chocolate were correctly recognized by 89-98% of the participants after completing the entire training program.

| Orthonasal versus retronasal odor perception
The participants were only trained and tested on orthonasal odor rec- developed for coffee professionals within the specialty coffee industry.Therefore the training of orthonasal odor recognition can be supported since an essential part of evaluating a specialty coffee is performed through smelling both the dry and the wet coffee grounds.
Nevertheless, the retronasal pathway represents more the actual drinking experience, and therefore it would be interesting to include retronasal odor tests when studying the odor perception of coffee.
Several studies show that odors are perceived differently whether they are experienced orthonasally or retronasally (Heilmann & Hummel, 2004;Hummel et al., 2006;Sun & Halpern, 2005).Increasing evidence supports that odors tend to be more intense when perceived through the orthonasal route than the retronasal (Heilmann & Hummel, 2004;Small, Gerber, Mak, & Hummel, 2005;Zhao, Scherer, Hajiloo, & Dalton, 2004).Consequently, the participants' improvement during this training program cannot be transferred directly to their odor recognition skills when drinking coffee.Nevertheless, adding a retronasal odor recognition test to this training program would be more complex, as it is difficult to avoid interference with other oral sensations such as gustation and thermal and mechanical stimulation in the mouth (Bojanowski & Hummel, 2012).For such a test, commercial test kits like the Flavor ActiV kit (GMP Flavor Standards), which requires capsules with specific aromas to dissolve in the specific beverage, could be considered as test and training tools.

| Learning effect
A part of the improvement in odor recognition may be due to a learning effect.Participants might have felt more confident at Test 2 and Test 3 compared with Test 1, which improved focus and attention on samples, influencing the response.As an example, this effect was seen in a study by de Lichtenberg, Broge, Wendin, Hyldig, and Bredie (2021) where an overall intensity score of two odors increased significantly from Test 1 to Test 2, without any training in between, although a 14 days wash-out period, that should be sufficient time for retesting, was applied (Nordin, Brämerson, Liden, & Bende, 1998).In this study, the time between each test day was 6 weeks, and as the control period indicates little but no significant increase in learning effect from Test 1 to Test 2, the learning effect is considered marginal.

| Benefits of the online sensory and cognitive training program
In Gave me more self-confidence when tasting coffee 9 1 2 7 9 The feedback received after each test day was useful to me  et al., 2014;Hummel et al., 2009;Konstantinidis, Tsakiropoulou, Bekiaridou, Kazantzidou, & Constantinidis, 2013).
This training program can be completed offsite in between panel training and sensory analysis, which can be beneficial for any sensory panel, as it requires less time and resources from the panel leader.Of the participants, 94% answered that the program made them aware of their strengths and weaknesses when tasting and smelling, whereas 97% stated that it was of good value for their sensory development; effects that would be beneficial for a panelist in any sensory panel.
The fact that 79% of the participants considered the training program fun is also a significant factor in keeping panelists motivated and engaged and should not be underestimated.For applications of the course to persons with smell losses, as seen in COVID-19 patients, adaptations to the course contents may be required.

| Brain and olfactory plasticity
This research builds on the increasing evidence that the brain structure can change in response to training.Magnetic resonance imaging (MRI) scanning's has observed structural changes in the brain in humans after many different kinds of skill acquisition and learning.For example, after intensive studying (Draganski et al., 2006), memory learning, and cognitive training (Engvig et al., 2010;Lövdén et al., 2010), after musical experiences (Gaser & Schlaug, 2003), foreign-language acquisition (Mårtensson et al., 2012), video game playing (Kühn, Gleich, Lorenz, Lindenberger, & Gallinat, 2014), after navigation (Maguire et al., 2000;Maguire, Woollett, & Spiers, 2006;Woollett & Maguire, 2011), and after the training of a new skill (Boyke, Driemeyer, Gaser, Buchel, & May, 2008;Draganski et al., 2004).In summary, many researchers have studied the macrostructural changes, mainly on gray matter, in the human brain after different learning experiences (Lövdén et al., 2013;Zatorre, Fields, & Johansen-Berg, 2012).The alterations have been reported already after 1 and 2 weeks of training (Driemeyer, Boyke, Gaser, Büchel, & May, 2008;Ilg et al., 2008) Wegener, Ilona, Antje, & Thomas, 2018).A study by Negoias, Pietsch and Hummel, 2016 showed that olfactory bulb volume increases in healthy people after 4 months of olfactory training, which build on the evidence that the olfactory system exhibits plasticity (Huart, Rombaux, & Hummel, 2013;Hummel, Stupka, Haehner, & Poletti, 2018;Negoias et al., 2010), and provides purpose for olfactory training.In general, the field of cognitive training is controversial.There is evidence that cognitive training enhances learning (Engvig et al., 2010;Lövdén et al., 2010), but researchers still discuss whether it transfers to other skills.Studies show that it is possible to improve on the task trained, and minor near transfer effects to other skills have been found on outcomes similar in type to the tasks used in the training (Karbach & Verhaeghen, 2014;Sala et al., 2019)  Other studies have demonstrated that statements on brain improvement during recruitment of participants increased transfer effects compared with participants recruited without focus on improvement (Green et al., 2019;Jaeggi, Buschkuehl, Shah, & Jonides, 2013).Since Fourthly, another fundamental principle of effective learning is feedback (Seitz & Dinse, 2006).Therefore, the participants received a comprehensive written feedback report after each test day with details on their performance to increase their learning.When asked, 85% of the participants found the feedback helpful.
Finally, a reward can be a significant motivational factor for learning.When the training program is designed to induce a certain amount of physiological excitement, it may maximize the training outcome (Green, Pouget, & Bavelier, 2010).This can, for example, be done by implementing game-like elements, such as a points system, game-like feedback, and game-like themes and animations, that can make the training more fun and engaging (Deveau et al., 2014;Mohammed et al., 2017).This was implemented in Task 3, which was built like a game, and in Task 5, where participants got to know their correct answer score.However, adding even more game-like elements in future training programs would be beneficial.
sensory and cognitive training program on improvement in odor recognition.The program was tailored to professionals working in the coffee business who wanted to improve their sensory abilities.The Le Nez du Café aroma kit ( Editions Jean Lenoir, Cassis, France) was chosen as a test kit for the olfactory performance training.This sensory kit is most commonly used for training and assessing coffee professionals' odor recognition ability, both in the SCA sensory skills courses and the Quality-grader course held by the Coffee Quality Institute (CQI).
study the effect of online sensory and cognitive training, a 12-week crossover trial was established.The participants joined two experimental periods, herein intervention and control.Before and after each period, the participants joined an onsite test for assessing their sensory skills.The participants received either intervention or control during the first period and switched after Test 2, as illustrated in Figure 1.The group that received the intervention in the first period is called the Intervention-first group, and the group that received the control period as the first period is called the Control-first group.The Intervention-first group (n = 28) was larger than the Control-first group (n = 16).To create relevant comparison results, the individuals in the Intervention-first group who were the best demographic match for the Control-first individuals were selected and remained in the Intervention-first group (n = 16).The residual in the Intervention-first group, not selected to be compared with the Control-first group, were transferred to the Intervention-first-residual group (n = 12).Other than improving the data relevance of any comparisons between the Intervention-first and the Control-first group, the Intervention-firstresidual group was analyzed for training effect between intervention and control period to test if the results could be repeated and generalized.2.2.2 | Intervention training: Daily training tasks The intervention consisted of a 6-week program during which the participants followed online daily sensory and cognitive training.The primary outcome measure was correctly recognizing the Le Nez du Café aroma set tested on each onsite test day.In this experiment, participants trained on the Le Nez du Café aroma set and several classical cognitive training tasks by daily accessing an online e-learning platform (developed by CFB-IT, Denmark).On the test day prior to the intervention training, the participants received a small training kit with the microtubes containing 36 different Le Nez du Café aromas, 1 per training day, and a small note per week with ideas for extra optional sensory and brain training exercises.The aromas were presented in coded 1.5 mL microtubes (polypropylene with attached lid, open diameter 1.08 cm, Sarstedt, Hounissen, Denmark), one for each training day, and coded accordingly.An 1 cm of a (0.5 Â 1.5 cm) coffee filter strip was dipped into the respective Le Nez du Café aroma bottle and placed into the coded microtube with a pincer.The microtubes were stored in the fridge (2-5 C) with the lids on for a maximum of 1 week prior to handing the training kit out to the participants.The cognitive training tasks were designed to train the participant's short-term memory, working memory, long-term memory, and ability to concentrate and stay focused.Each day the participants had six tasks to complete.An overview of the daily training tasks during the intervention program is seen in Table2.The first task was to smell and identify the odor of the day.Task 2 was a squared breathing exercise that revealed the correct answer to Task 1 by showing a picture of the ingredient from where the odor derives.A squared breathing exercise is a technique where participants take slow, deep breaths.The participants had to look at the picture and imagine the smell of the aroma every time they breathed T A B L E 1 Comparison of baseline characteristics between the three groups of the total of 44 completing participants.

F
I G U R E 1 Experimental study design.The figure shows the three groups that followed the training program, and the crossover between intervention training and active control period.The dashed line shows the Intervention-first-residual group that was added to test the reliability of the main results.T A B L E 2 Overview of daily training tasks during the intervention training.remember the aromas from Task 4? Long-term memory taste and smell.They completed several identification, ranking, and threshold tests with the five basic tastes and a coffee-tasting session, although this paper focuses on the odor recognition test results.On each test day, the participants had to identify the 36 Le Nez du Cafe aromas.The test was conducted at the same time of the day for all three test days to reduce noise from any daytime variation in olfactory sensitivity.The participants were presented with the aromas in three sets of 30 min duration (12 aromas per set), with basic taste identification-, ranking-and threshold tests and 30 min break in between.All participants were instructed on the test details by a trained experimenter, and identical instructions were repeated on Test 2 and Test 3. The participants were instructed only to sniff the Le Nez du Café aromas once to avoid odor fatigue, although they were allowed to re-sniff once if necessary.To avoid odor fatigue and adaptation, a 30 s break was applied between each sample.All test sessions were performed in quiet and well-ventilated rooms under the responsibility of the trained experimenter.The odor recognition test was designed as a 12 alternatives force choice identification test.For each aroma smelled, the participant was asked to identify the name of the odor from 12 alternative options as presented in Table 4.The 12 options were presented before the participant had smelled the odor.The alternative options were selected among the Le Nez du Café aromas used in the test.The test results were not presented during the test session, but each participant received their results in a feedback report the day after the test.The Le Nez du Café aromas were presented in randomly coded brown bottles (soda glass, DIN 32, 30 mL, open diameter of 4 cm) with screw-top (DIN32, PP, with LDPE seal; Buch & Holm, Denmark).
To analyze the difference in odor recognition score before and after the intervention training, both for each odor individually and the participants in each experimental group, paired-sample t tests were performed.To explore the possible linear correlation between the amount of training and improvement in odor recognition score Pearson's correlations (two-tailed) was conducted based on the normally distributed data.All statistical analyses were evaluated based on a significance level set at p < .05.Statistical analyses were performed with IBM SPSS Statistics version 25.
challenging aromas to detect were coffee pulp, roasted coffee, toast, and black currant (≤20% of the participants recognized them at Test 1).At Test 3, more than 90% of the participants detected four aromas: Dark chocolate, potato, clove, and lemon.While roasted coffee, medicinal, black currant, coriander seeds, rubber, and roasted almonds remained difficult for the majority as <50% had them correct after the 12 weeks of training.The aromas with the most remarkable improvement were coffee pulp, toast, roasted peanuts, garden peas, clove, cooked beef, straw, and pepper; they all increased by more than 30%.In Figure2, the mean percentages of correct odor recognition at baseline, after intervention training, and after the control period are shown.At baseline, all participants had 49% correct answers on average, and no significant difference between groups was observed.The Control-first group did not improve (p = .502)their odor recognition score from Test 1 at baseline to Test 2. Thereafter by following the intervention training program, they improved their odor recognition score significantly with 17% on average (p < .01)at Test 3. The Intervention-first group participants improved their odor recognition scores significantly after the intervention training at Test 2 (p < .01).On average, they had 12% more correct answers at Test 2 compared with Test 1.After 6 weeks of the control period, the Intervention-first group maintained the level of odor recognition learned.No significant difference was observed between Test 2 and Test 3.The Intervention-first-residual group, formed to check for generalizability, showed exactly the same pattern as the Intervention-first group.The participants improved their odor recognition scores significantly (p < .001).On average, they had 16% more correct answers at Test 2 compared with Test 1.After 6 weeks of the control period, the Intervention-first-residual group also maintained the level of odor recognition learned.No significant difference was observed between Test 2 and Test 3.
presented and show a pattern of more likely choices of training combinations.On 46% of the days, all of the six training tasks were F I G U R E 2 Mean percentage of correct odor recognition at Test 1, Test 2 and Test -3 for the Controlfirst group (n = 16), the Interventionfirst group (n = 16) and the Intervention-first-residual (n = 12) group, respectively.total), 44 participants Â 41 training days = 1804 training days; n, number of training days; %, percent of the 1804 days when this combination was completed.
ment (p < .001)and a correlation with a lower significance between days with any training completed and improvement (p = .019).The r values indicate that an increase in the number of days with all training completed and in the number of days with any training completed both increased the likelihood of improvement in odor recognition; therefore, the more tasks completed, the more remarkable improvement occurred.This emphasized the possible importance of cognitive training on improvement.3.2 | Participants' experience with the sensory and cognitive training programAfter completing the sensory and cognitive training program, a questionnaire was sent to the 44 participants who completed the entire program.In total, 34 (77%) participants filled out the questionnaire (Table9).Overall, the participants were very satisfied with the training program.More than 85% agreed that the course improved their ability to detect not only the Le Nez du café aromas but aromas in general and also describe the aroma of coffee.The improvement in cognitive skills was a little less; nevertheless, more than 70% agreed that they improved their memory and ability to stay concentrated and pay attention.It is important to mention that the course made the participants aware of their strengths and weaknesses when tasting and smelling, and 79% felt that the course gave them more selfconfidence when tasting coffee, whereas 97% agreed that the course was of good value for their sensory development.Finally, 79% of the participants found the sensory training course fun to take part in. 4 | DISCUSSION This study aimed to develop an online sensory and cognitive training program tailored for coffee professionals to improve their ability to recognize coffee odors.The findings of the research demonstrate a positive and statistically significant effect of the training program on olfactory improvement.The participating coffee professionals improved their odor recognition score of the 36 Le Nez du Café aromas after only 6 weeks of training.
the times of the coronavirus disease 2019 (COVID-19) pandemic, online teaching and training have shown to be even more important than when this training program initially was developed.At the same time, the need for and interest in olfactory training programs for people with smell loss due to the virus infection have dramatically T A B L E 9 Participants feedback after the course rated on a 5-point Likert scale.
disagree and strongly disagree.b The sum of agree and strongly agree increased.Several studies support that olfactory training can improve general olfactory function in patients with smell loss (Damm . Therefore, the 6 weeks duration of the training program in this study may be sufficient for change in brain function, although studies on olfactory training have shown effects after longer periods of training (3-6 months; Oleszkiewicz, Abriat, Doelz, Azema, & Hummel, 2021;

4. 6 |
Transfer effect: Cognitive training improves odor recognition The online training program is a novel tool for improving the odor recognition of coffee professionals.The combination of olfactory and cognitive training seems promising, although the impact of the cognitive training alone is uncertain.

4. 7
| Motivational factors Different motivational factors may have influenced the outcome of the cognitive training program, including the participants' expectations, the duration, and spacing of the training, the level of difficulty, their comprehension of the feedback, and rewards.Firstly, the participants' expectations that the training program was designed to improve their sensory performance might have led to a beneficial outcome of the training, similar to the well-known placebo effect.Before attending the sensory performance course, the participants knew that the researchers hypothesized that cognitive training can improve sensory performance.That knowledge may have had a positive influence on the transfer effect and thereby test results.
this study was promoted as a sensory performance course, expectations were created intentionally for both the intervention training and Sensory Studies 1745459x, 2023, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/joss.12819 by Det Kongelige, Wiley Online Library on [21/06/2023].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License the control period and could therefore possibly have led to increased success of not only the intervention, but the whole training program.Therefore, the expectation effect can be considered equal for both the intervention training and the control period.Second, the optimal duration and spacing of training are critical motivational factors.Pearson's correlation analysis shows that the number of training tasks completed correlated significantly with improvement in odor recognition score.The more the participants trained, the more they improved.This corresponds well with the idea that the amount learned depends on the time spent learning, demonstrated by several researchers within cognitive training (Jaeggi, Buschkuehl, Jonides, & Perrig, 2008; Schmiedek, Lövdén, & Lindenberger, 2010; Stafford & Dewar, 2014).The total amount of time spent on training is essential, but the training sessions should be distributed over time for the most efficient learning outcome (Baddeley & Longman, 1978).The training program of this research was developed with many daily short-training sessions instead of fewer, more extended sessions.Whether the length of the training sessions and amount of training days applied in this study are optimal is uncertain.Many participants did not complete the last training days; nevertheless, the positive effect of the training was still observed, and therefore fewer days of training may be considered in future studies.Third, adjustment to the training tasks is important to consider.The level of difficulty of the training tasks increased each week, intending to reach a level of difficulty that kept the participants engaged.Increasing the difficulty level when participants become proficient at completing a training task is known to be an effective learning tool (Deveau, Jaeggi, Zordan, Phung, & Seitz, 2014).However, it would be more beneficial to design the program with individual dynamic difficulty adjustments.When a participant becomes proficient at completing a training task, the task's difficulty level should be increased for that participant particularly.In this research, the last 2 weeks' training tasks may have reached a level of difficulty too advanced for some participants who lost the motivation to continue training.

4. 8 |
Limitations/perspectives The online sensory training program was developed for coffee tasters and showed merits for their sensory performance skills.Developing the program for other target groups would require adjustments.The present training involved orthonasal training, and further studies are required to evaluate whether these skills are transferrable to performing taste evaluations.This work included training of cognitive and sensory skills.Further research should investigate whether cognitive training alone could improve odor recognition skills and thus reduce the complexity of the training program.The training program developed offers the possibility to investigate how an online training setting can help a person to improve his or her sensory skills.This is valuable not only for coffee professionals but for any sensory panel, regardless of the food substance, mainly because it does not require much time or many resources from the facilitators' side.5 | CONCLUSION In summary, this study shows promising results for online sensory and cognitive training supporting the learning of olfactory recognition.The test and training method aimed to improve participants' odor recognition score of the 36 Le Nez du café aromas.On average, the participants improved their odor recognition score by 15% after 6 weeks of training (p < .01).However, the improvement scores of each specific aroma ranked from no improvement to 39% improvement in odor recognition depending on the familiarity and difficulty of each aroma.A positive correlation between completed training and percentage improvement was observed (p < .001).The participants enjoyed the training as they found it fun and beneficial for sensory improvement.The program shows potential as a new training tool for improving odor recognition in coffee professionals.It is a new and promising approach to perform more sensory training online.However, more work is needed to understand the effect of cognitive training and to expand the results to other groups than coffee professionals.ACKNOWLEDGMENTS This study is a part of an industrial PhD project supported by a grant (5189-00086B) from the Innovation Fund Denmark and CoffeeMind.Without partner organizations worldwide, this project would not have been possible; the time, effort, resources, and risks taken by Coffee Support, Rose Van Asten, Acts29, and Arabian Coffee Institute to make this happen are deeply appreciated.Jesper Alstrup, Christina Josefine Birke Rune, Signe Gadegaard, Christian Behrens, and Konstantinos Arapogiannis are thanked for their contribution to developing the sensory training program.
Example of response alternatives selected for each of the 12 Le Nez du café aromas in one of the three test rounds.Percent of participants with correct odor recognition at Test 1 and 3.Note: Δ improvement and one-sided significance level of the 36 Le Nez du Café aromas.Aromas sorted according to Δ improvement.
a See Appendix A. Sensory Studies 1745459x, 2023, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/joss.12819 by Det Kongelige, Wiley Online Library on [21/06/2023].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)onWiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License T A B L E 4 Alternative options for the 12-alternative forced choice test.1745459x,2023, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/joss.12819 by Det Kongelige, Wiley Online Library on [21/06/2023].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License group.This intervention-first-residual group consisted of 12 participants also started with the intervention training.In Table 1, baseline characteristics of the three groups are shown: the Control-first group, the Intervention-first group, and Intervention-first-residual group.The Control-first and Intervention-first groups were similar in gender distribution and age.respectively.After the 12 weeks training program, the recognition of 20 aromas improved significantly, while no improvement was seen for the aromas rubber and lemon.Lemon was the only aroma recognized by all participants already at Test 1; no improvement was possible.At Test 1, more than 70% of the participants already had licorice, potato, cucumber, dark chocolate, and lemon correct, while the most T A B L E 5 1745459x, 2023, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/joss.12819 by Det Kongelige, Wiley Online Library on [21/06/2023].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Type of training tasks completed.
ognition and not on retronasal perception.The training program is T A B L E 7 *Correlation is significant at the .05level(two-tailed).**Correlation is significant at the .01level(two-tailed).1745459x,2023, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/joss.12819 by Det Kongelige, Wiley Online Library on [21/06/2023].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License