*3.2. Consumer Liking of Innovative Food*

The evaluation of dishes is a complex experience that includes sensory, emotional and ideational components. For this reason, the experienced and expected-liking and the influence of factors associated with consumer attitudes toward new food were explored.

*Hedonic Image.* Expected and experienced-liking of traditional and molecular versions of lemon- and tomato-based dishes were not varied (Figure 3). However, the expectedliking of NbN in both meals was significantly higher than the experienced-liking.

**Figure 3.** Expected and experienced overall liking in relation to dishes: a, b, c—values of expectedliking or experienced-liking that are marked by different lowercase letters between various versions of dishes with the same main ingredients differ significantly at *p* ≤ 0.05; \* asterisk brackets indicate a significant difference between expected and experienced overall liking of the given sample at *p* ≤ 0.05.

The highest positive correlations were observed between the expected-liking and appearance-liking (rho = [0.33–0.73]; *p* ≤ 0.05). In turn, the experienced-liking depended significantly on flavor-liking (rho = [0.82–0.88]; *p* ≤ 0.05) and texture-liking (rho = [0.61– 0.80]; *p* ≤ 0.05). Generally, the liking based on appearance of NbN dishes had the lowest effect on experienced liking.

All the dishes received moderately high hedonic scores; however, their deeper examination showed differences in the degree of liking, depending on type of dish and their version. The hedonic scores for the experienced-liking of the Note by Note dishes were lower than for traditional versions (*p* ≤ 0.05) but did not vary from scores of molecular courses (Figure 4). The odor, taste/flavor, and texture-liking for traditional meals were significantly higher (*p* ≤ 0.05) than molecular and NbN versions of lemon-based meals and the NbN version of tomato meals (Table 3).



a, b, c—mean values marked by different letters in verses, differ significantly at *p* ≤ 0.05.

**Figure 4.** The dishes and term representation in the first and second coordinates of the Correspondence Analysis performed on the frequency of use of attributes by consumers (according to check-all-that-apply).

*Factors influencing liking.* The higher standard error of the mean hedonic scores exhibits varied affective response of consumers (Table 3, Figure 3). The effect of food neophobia and consumer innovativeness' levels were evaluated. The explanation of the results might be found in the Food Neophobia Level (Table 2). Higher levels of Food Neophobia (FNL) of evaluators resulted in less liking for texture (r = −0.27, *p* ≤ 0.05) for the molecular cookie. Moreover, the higher FNL, the lower the experienced-liking for the NbN cookie (r = −37, *p* ≤ 0.05). This phenomenon was especially evident in the group of the most neophobic participants. The difference between mean values of expected and experienced-liking was the highest (2.0–4.7 c.u.) in the group of the most neophobic subjects, while in the other groups (the most neutral and the most neophilic participants), it ranged between 0.4–2.2 c.u.

As consumer innovativeness level increased, the flavor-liking (r = −0.36, *p* ≤ 0.05) and experienced-liking (r = −0.40, *p* ≤ 0.05) of traditional lemon cookie decreased. The familiarity with novel cuisines did not show any significant effect on the liking.

*Willingness to try by consumers innovative foods in the future.* A higher percentage of consumers showed a willingness to try similar traditional dishes (71.9–86.0%) in the future, than the molecular (54.4–61.4%) or Note by Note (59.6–64.9%) ones (not shown in tables). However, the percentage of people who did not want to try molecular (17.5–31.6%) or NbN (17.5–19.3%) versions was similar to that of the undecided consumers (15.8–21.1% and 17.5–22.8%, respectively). The mean values of liking in a group of consumers who would not eat modernist dishes again were lower than 4.5 c.u. The mean score of undecided consumers (5.1–6.5 c.u.) was lower than those who declared a willingness to eat again (6.5– 6.8 c.u.). This may indicate that the consumption of dishes prepared with these innovative techniques introduces a dissonance between unfulfilled sensory expectations and the desire to be surprised.

#### *3.3. Emotions and Sensory-Hedonic Sensations of Meals*

The sum of word expressions indicated by consumers used for the survey (*n* = 56) revealed (data not presented) that molecular and Note by Note dishes evoked more emotions than traditional versions. The NbN cuisine aroused the most emotions (366–405 indications in total), followed by molecular cuisine (300–330) and the least traditional cuisine (303–308).

Data analysis further showed that traditional cuisine compared to modernist ones were associated with word expressions such as tasty (43 indications per meal), traditional (35) and typical (31). The proportions of selections by consumers for each attribute of CATA question for all dishes was presented in a contingency table (Table 4).

**Table 4.** A contingency table of data set with 6 products and 22 emotional attributes of the CATA question.


Cochran's Q test was carried out to determine whether the proportions of selection by consumers for each attribute of the CATA question differed, taking into account the assessed dishes. Post hoc multiple pairwise comparisons were carried out using McNemar's test with Bonferroni alpha adjustment. Different superscript letters within each row denote significant differences (*p* ≤ 0.05).

> A higher proportion, e.g., closer to 1.00, indicates that, among the six dishes, the attribute was more frequently chosen by consumers. Cochran's Q test revealed differences for 19 of the 22 terms of the CATA questions used to characterize the dishes. The dishes did not vary in attributes such as dietetic, artificial, and fresh. As the use of processed ingredients increased, the number of the following indications increased: surprising, intriguing, innovative, atypical, unique and trendy. These relationships can be clearly seen on a bi-plot of correspondence analysis (explaining 91.4% of total variance), which show associations between the type of dishes (culinary techniques) and emotional descriptors or hedonic sensations. Along with the level of food processing, e.g., for lemon-based dishes (Classic → Molecular, Note by Note cuisine), the disappointment increased when the naturalness sensation decreased (Figure 4). Interestingly, a larger share of consumers perceived NbN dishes as slightly more unique, trendy and innovative than molecular ones.

#### *3.4. Mimic Expressions Assessment: Type and Level of Emotions in Relation to Dishes in Time*

Consumption of traditional meals aroused fewer facial expressions among consumers than molecular or Note by Note ones. In the case of traditional courses, the intensity of neutral facial expression ranged between 0.38 and 0.58 c.u., while in molecular and NbN dishes were lower (respectively, 0.28–0.53 and 0.26–0.53 c.u.). The level of neutrality depended on the moment of consumption. It was at a stable level in traditional dishes, while in the case of modernist cuisines, it had different dynamics in time. The consumption of molecular and Note by Note dishes evoked fewer neutral facial expressions while

chewing, which rose significantly after swallowing samples. Neutrality rose 4.9–9.2% in traditional versions, 16.6-24.5% in molecular cuisine and 29.1–33.4% in NbN.

Mean values from facial expression measurement (*n* = 15) and mean values of experiencedliking (*n* = 56) revealed a significant positive correlation (*p* ≤ 0.05). Higher intensity of neutrality (only during the first 10s of measurement) and lower intensity of disgust in 15s of assessment (average deglutition time) resulted in higher scores for experienced-liking (respectively r= 0.82–0.92 and r = 0.96).

Analysis of facial expressions during the entire consumption time revealed that the dominant expression (excluding neutrality) associated with the eating of traditional lemon cookie and tomato soup was happiness (Figure 5a,c). Although kept at a relatively constant level, the emotion was more intense than other emotions from 5 to 20 s after placing the sample in the mouth and from 35 s until the end of measurement.

**Figure 5.** Real-time mimic expresions intensity towards dishes based on (**a**,**b**) lemon and (**c**,**d**) tomato: (**a**)—during entire consumption time of lemon-based dishes; (**b**)—for the swallowing-related time of lemon-based dishes; (**c**)—during entire consumption time of tomato-based dishes; (**d**)—for the swallowing-related time of tomato-based dishes.

In the case of molecular and NbN versions, quite different dynamic changes in expressions can be observed. Moreover, facial expression status during consumption depended on the type of dish. The common feature for both of the lemon-based products examined is that the highest peaks of happiness appear around 20 s from placing the sample in the mouth and then disappear (Figure 5a). Simultaneously, moderately high intensities of surprise mixed with disgust are exhibited during the entire consumption.

Consumers reacted differently for the tomato puffy rice snack. Its consumption did not evoke any happiness, and during the first 10 s of FaceReader measurement, a high level of surprise rose and then decreased (Figure 5c). After 15 s of measurement, the emotions decreased, probably as the result of snack texture. Mimic expressions assessment of the rice soufflé with tomato gel revealed that the dominant expression prior to 15 s of measurement was disgust, followed by happiness. Afterwards, notable peaks of surprise and happiness were recorded. Some facial mimic expression may be related to the sensory properties of dishes (Figure 3) and the feeling of aftertaste in time, both positive and negative.

The analysis of facial expressions dynamics for swallowing-related time (Figure 5b,d) revealed that prior to the swallowing (−5 s) of both traditional dishes, slight surprise

peaks were noticed, followed by a neutrality decrease (+5 s). In the case of molecular and NbN lemon cookies, notable (especially in the molecular version) peaks of happiness were noticed 5 s before swallowing. At 5 s before the swallowing of molecular and NbN versions of tomato soup, a slight facial expression of disgust was noticed. In all molecular and NbN versions, a growing trend of neutrality was observed, which was opposite from classical ones.

#### **4. Discussion**

The detailed sensory profile is a very important tool for new product development. It allows one to explore the effects of ingredients and processing variables on the final sensory quality of foods and dishes. Moreover, such information is very useful for understanding consumer responses in relation to products' acceptance. The present study revealed that each version of the lemon and tomato dishes (traditional, molecular, NbN) had a separate sensory pattern (profile) related to technological aspects of their production and the recipe used. It is crucial for cooks and producers to maintain a balanced intensity of attributes in various meals that determine their identity and specificity regardless of the modernization process. Thus, too intense or weak intensity of, for example, key flavor descriptors and basic tastes like sourness or sweetness can result in a hedonic response in relation to a product. That, in turn, can potentially affect the degree of liking by consumers and their positive or negative experiences, including evoked emotions [65]. According to Spinelli and Jaeger [66], the sensory characteristics of a product may be altered in order to increase specific positive emotions or even to decrease negative emotions among consumers. This seems crucial in modern cuisine and it is also intriguing due to the atypicality of the dishes.

Visual appeal of the product is an important driver of sensory expectations, which was confirmed in the present study by the highest correlation of expected-liking and appearance-liking. The experienced-liking of NbN dishes was lower than that of traditional ones. The most important factor for consumers was the flavor/taste of the product. These results are in accordance with findings of Santagiuliana et al. [67], who stated that the visual appearance of novel heterogeneous foods can significantly affect expected-liking but not the actual-liking. The experienced-liking is affected by textural and flavor oral sensory perceptions. However, visual cues can significantly influence consumer texture perception. Moreover, literature [68] confirmed that consumers primarily pay attention to flavor-liking when evaluating the overall-liking of products. The examined meals, depending on the type, had a completely different appearance, which could determine the perception of taste and flavor. It is intriguing the extent to which the consonance of the color with the product (yellow lemon cookies) affected the sensory perception of consumers in contrast to the incongruent color of dishes in the case of tomato soup (e.g., white NbN version). Moreover, different textures of the dishes could affect the results of the assessment. Liking of evaluated dishes may be related to their sensory profile. For example, low intensity of buttery attributes as well as a higher level of lemon note in the molecular cookie and artificial lemon attributes (odor and flavor) in NbN compared to the traditional products could determine a lower hedonic response of consumers. Similarly, the significantly lower intensity of tomato flavor and aroma in modern cuisine (molecular, NbN) probably affected the lower results of liking compared to the traditional option. According to other studies [69,70], change of texture and appearance of familiar food in novel food development are only possible when similar retro-nasal sensation is ensured. Consumers are sensitive to the modification of the expected texture. Traynor et al. [48] emphasize that besides intrinsic sensory properties, extrinsic motivation like emotional reactions have a significant effect on the acceptance of novel foods.

Consumer expectations regarding odor, taste, and texture-liking of modernized versions of dishes were probably higher. Similar results were demonstrated in a study by Stolzenbach et al. [71], where the mean scores for liking were significantly higher for the traditional honey than for the novel kind. Moderately high sensory ratings of the molecular version of Portuguese custard tart 'Pastel de Nata' were stated by Oliveira et al. [72]. The

lowest appearance-liking for the molecular version of lemon cookies may be the result of food looking more chaotic (cookie chunk stuck into sherbet, which is served in soup). On the other hand, a congruent color with a product representing an unusual shape influenced a significantly higher expected overall-liking when compared to the tomato version (white soufflé soup color, NbN version). This is in line with the research of other authors. Zellner et al. [73] and Zellner et al. [74] provided scientific evidence that foods presented in a less neat manner are less liked. Similar to our findings, results of studies with novel salad dressings [75,76] and with other novel foods [77] indicated a higher mean of expected acceptability in a group of neophiliacs than neophobics.

According to the psychological theories, there are differences in consumer's product expectations [51]. Piqueras-Fiszman and Spence [51] stated that disconfirmation (to any degree) between a consumer's product expectations and the experience can be explained by the two phenomena. The assimilation is one of them, and it happens when the consumer tries to minimize differences between the perception of the product and its expectation, while the contrast happens when the consumer tries to maximize these differences. In our study, consumer attitudes toward examined meals include these two aspects. The assimilation probably occurred for the first two methods (traditional and molecular meals), while the contrast appeared for NbN. This suggests that expectations of consumers were set too high, which resulted in a certain level of dissatisfaction. This also may indicate that the consumption of dishes prepared with these innovative techniques creates dissonance between unfulfilled sensory expectations and the desire to be surprised by novelties. It is worth mentioning that small deviations from the level of consumer adaptation can be seen as interesting and novel. Large deviations tend to result in disgust or neophobia instead [51,54]. It is worth emphasizing that the research results could be somewhat different if the meals were assessed in a different surrounding, for example, in a restaurant. According to many authors [31,44,78] consumers may be more willing to tolerate or expect a certain level of incongruency in some contexts (e.g., modernist/experimental restaurants).

CATA questions provided more profound insight and better explain consumers' perceptions and liking of modern cuisines (molecular and NbN versions of dishes). Traditional cuisine compared to modernist ones were associated with being tasty, traditional, and typical. In contrast, novel cuisines we more frequently perceived as surprising, intriguing and innovative. Our results appear to be well substantiated by the results of several authors, who compared traditional foods with their novel versions. Consumers perceived novel beers as "unusual", "complex", "intriguing" and for "special occasions" [79]; molecular dishes as "surprising", "arousing curiosity", "innovative" and "liked" [41]. Moreover, a familiar type of chocolate was associated with positive emotion terms (e.g., sweet—"happy"), but new chocolates (e.g., with salt) evoked various emotions, including "bored", "interested", or "a little naughty" [80]. Loss et al. [81] also concluded that the acceptance of very innovative foods go beyond neophobia. The results of previous studies [81,82] reveal that traditional versions of selected foods were more liked in comparison to their modernized versions. Additionally, the most unusual foods, arousing curiosity and surprise and challenging for senses, were least liked. Acceptance of unfamiliar food is mainly dependent on its degree of similarity to familiar food [44], which aligns to some extent with the results of our research. In this context, it should be emphasized that the unique, aesthetic and atypical appearance of new products should also correspond with the experienced and balanced taste and flavor that provide satisfaction, evoke positive emotions and affect the degree of liking. It seems that this will be met in particular when the expected-liking is relatively high, but equal to the experienced-liking or even lower. In a previous study [41], modernist cuisine also aroused more positive emotions (35.3%) in consumers than the traditional one.

Mimic expressions assessment by FaceReader was a measure to capture the type and level of emotions as temporal effects in relation to examined meals. According to Danner and Duerrschmid [83], the unfolding of emotions can provide deeper insight into the evaluation process itself and the formation of liking. Our study revealed that higher neutrality during 10 s and disgust in 15 s of consumption was related with higher experienced-liking. These findings align with the result of Danner et al. [84], who found a high positive correlation between "neutral" facial expression with liking, as well as a high negative correlation of "angry" and "disgusted" facial expression with liking. Neutral emotion is frequently used to describe hedonically liked stimuli [85]. In He et al. [42] study, pleasant odors were related to more intense neutral and surprised facial expressions as well as less intense disgust emotion. On the other hand, unpleasant odors evoked fewer neutral reactions and a more intensified disgust. Generally, a higher concentration of odors resulted in lower facial neutrality and more scared expressions. It should be noted that fewer consumers (*n* = 15) agreed and participated in the FaceReader test than in other tests performed with consumers. According to Crist et al. [86], a range of 10– 50 participants would take part in the study using automated facial expression analysis software. They found that the number of participants would vary depending on flavors, flavors intensity and expected treatment acceptability. It was emphasized that products with smaller flavor differences require more participants. Our dishes varied significantly in sensory characteristics (profiling data). Nevertheless, the results of the research with FaceReader should be considered as preliminary.

More consumers would like to eat similar traditional dishes in the future than molecular or NbN dishes. Consumers were a little disappointed by NbN. This was a result of the unfamiliar appearance of dishes, especially the tomato-based dish. The lemon-based dish in NbN version was similar to a dessert, and it affected the evaluation. According to Burke et al. [87], making the NbN dish that looks familiar to the consumers may help to overcome their neophobic reaction. The inspiration from traditional foods may contribute to the success of NbN cooking and cuisine. In previous studies, consumers declared moderately high willingness to buy the molecular vinaigrette jellies (42%), a port wine faux caviar (58%) or powdered olive oil with flavors (72%) [2,45,46]. Most of the participants (88%) in our recent study [41] would like to taste different molecular courses in the future. In turn, consumers in the study by Mielby and Frøst [44] did not want to eat the most innovative and highly unusual molecular dishes again. This is supported by the research of Oliveira et al. [72], wherein purchase intention was moderately high. Tan et al. [88] have also reported that willingness to eat novel burgers with insect additions was lower than regular beef patties. Their findings suggest that consumers may be inclined to taste novel (unusual) foods to satisfy their curiosity, but not to consume them again, especially in the case of low sensory-liking or inappropriateness for consumption. The results of Barenna and Sánchez's [36] study suggest that decisions to consume novel foods have a more pronounced emotional component in a group of neophobic consumers. Hence, neophiliacs try novel foods in restaurants more often than the neophobes [89]. Neophiliacs had a higher awareness, willingness to try and rated unfamiliar food more favorably [75].

The detailed sensory profile is a very important tool for new product development. It allows us to explore the effects of ingredients and processing variables on the final sensory quality of foods and dishes. Moreover, such information is very useful for understanding consumer responses to modernist products such as molecular and NbN dishes. Higher results of expected-liking compared to Note by Note experienced-liking showed disappointment with this version which may be related to the flavor and taste of the dishes. Additionally, significantly lower overall liking of NbN dishes compared to classic ones suggests that the success of novel dishes might be related to the other elements of multisensory experience. This type of modern cuisine is not designed for a large audience. It seems to be intended for small groups of consumers, with low neophobia and who are very familiar with modern cuisine. This finding requires further research.

#### **5. Conclusions**

Our findings show that the expected-liking of dishes depended on their type. The results may be related to appearance and especially color–taste congruency with the dishes, which affect consumer associations and expectations regarding sensory characteristics. Moderately high liking scores of molecular and Note by Note (NbN) cuisines combined with the structure of the emotional response suggests the new insight that consumption of such dishes is as result of temporary emotional arousal derived from an element of novelty. Consumption of modernist dishes evoked more mimic expressions among participants than traditional ones, especially during the first phase (chewing). This facial emotional arousal was confirmed by a number of declared emotions and hedonic sensations. The success of modernist dishes that are served at high-end restaurants is not only strongly associated with acceptable intrinsic sensory characteristics but also with enhanced emotional reactions.

On the other hand, it is worth emphasizing that food neophobia and consumer innovativeness had significant effects on the liking of given products and affected the attitude of consumers towards modern dishes and their willingness to try them again. A higher percentage of consumers showed a willingness to try similar traditional dishes in the future than molecular or NbN dishes, with more than half of the consumers indicating they would eat similar modernist dishes. This may indicate that the consumption of dishes prepared with these innovative techniques introduces a dissonance between unfulfilled sensory expectations and the desire to be surprised by novelties.

Traditional cuisine was associated with a greater number of consumers' word expressions such as tasty, traditional and typical compared to modern ones. A higher level of food processing (molecular cuisine → Note by Note cuisine), caused more participants to perceive meals as surprising, intriguing/interesting and innovative. The uniqueness of modern dishes was most related to innovativeness and being surprising and intriguing, while disappointment in modernist dishes may have been associated with a lower sensation of naturalness (e.g., lemon-based dishes) compared to classical ones.

New food product developers should take into account the attitudes and expectations of potential consumers. In the future, focus groups could be planned to identify different consumer expectations for new products. Such an approach would be valuable in modifying features like the taste, flavor and texture of dishes according to consumers' points of view. Culinary workshops for consumers combined with the preparation and evaluation of such kinds of dishes would be also interesting.

Learning about the perception of modernist cooking solutions enables a better understanding of the sensory experiences of potential consumers, which becomes the space and inspiration for innovation. According to these findings, and taking into account the fact that it is difficult to assess the type of customers trying the dishes (neophiliacs, neophobes, neutrals), we believe that more attention should be given to the appearance of new products. The acceptance of innovation is also linked to the risks of failure. If we change the appearance of the food too much, the exposure to failure will be increased.

#### **6. Limitation**

A limitation of our research is that assessments of hedonic and other tests were carried out under laboratory conditions, while the multisensory experience of modernist cuisine implies that it is associated with the restaurant atmosphere, including social contacts and service provided. Moreover, the evaluation of these innovative dishes was performed by comparison to traditional dishes, which could have had an impact on the results.

Application of the immersive approach would recognize contextual influences (e.g., external variables like colors, light, temperature) on perception of modernist dishes, simultaneously with consumer engagement. Such research would allow a deeper analysis and understanding of consumers behavior and expectations towards molecular cuisine and Note by Note cooking.

Another limitation of our research is the number of consumers taking part in the experiment. There were only 56 consumers that took part in the experiment. Amongst them, more than a half were classified as the most neutral, and about 20% were categorized as the most neophiliac, who are eager to try new products.

**Author Contributions:** Conceptualization, A.G. and E.C.-S.; methodology, A.G., E.C.-S. and E.K.; data curation, A.G., G.W.-Z. and K.B.; writing—original draft preparation, A.G., E.C.S. and E.K.; writing—review and editing, A.G., E.C.-S. and E.K.; visualization, A.G.; supervision, E.C.-S.; data analysis, A.G. and E.K. All authors have read and agreed to the published version of the manuscript.

**Funding:** Research financed by Polish Ministry of Science and Higher Education within funds of Institute of Human Nutrition Sciences, Warsaw University of Life Sciences (WULS) for scientific research.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Ethics Committee of the Faculty of Human Nutrition and Consumer Sciences, Warsaw University of Life Sciences–SGGW (Poland) on 22nd January 2019 (Resolution No. 03/2019).

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.Acknowledgments: Thanks to Andrea Camastra, the Head Chef for enabling Kacper Bylinka to prepare Note by Note dishes in Senses Restaurant in Warsaw, Poland.

**Conflicts of Interest:** The authors declare no conflict of interest.
