Sensory Characterization of Conventional and Diets in Sugar Guava Sweet by Check-All- That-Apply and Emotion Measurement Methods

This study aimed to investigate the sensory characterization of commercial diets in commercial guava sugar samples by the Check-all-that-apply (CATA) and Emotion Measurement methods. The samples were evaluated by 120 consumers using the CATA method, (CATA) Ideal, Acceptance Test, Purchase Intention, Emotion Measure, and Ideal Test in relation to the sweetness and taste of guava, allowing the construction of two External Preference Maps aimed to associate the results from the overall impression with the terms from CATA and the emotional responses. The samples of conventional C1, C2, and C4 showed greater correlations with the descriptors terms and positive emotions, and the samples of diets D2 and D4 showed greater correlations with the descriptors terms and negative emotions, which was also observed in the acceptance and purchase intention tests, showing the consumers preference for the product with high caloric value.

Introduction

Guava, fruit of the guava tree (Psidium guajava L.), besides being consumed in natura, it is used in the food industry to obtain products such as juices, nectars, pulps, ice cream, sweets, jams, jellies, and others [1]. The sweet in paste is defined as the product of proper processing of disintegrated edible parts of vegetables with sugars, with or without the addition of water, pectin, pH adjuster, and other ingredients and additives allowed until an appropriate consistency is obtained [2].

Diet foods are intended for diets with nutrient restrictions such as carbohydrates, fats, proteins, and sodium, among others, so that their composition meets the needs of people under specific metabolic, physiological, or pathological conditions [3]. They are generally targeted to people who have some disorder in the sugar metabolism (diabetics) or consumers who demand of low caloric food products [4]. In the last decades, dietary product consumption has been enhanced due to the increase of chronic nontransmissible diseases, especially obesity and diabetes mellitus [5].

Sensory evaluation measures the reaction to stimuli resulting from product use or consumption through affective tests [6]. The sensory evaluations that use word association methods bring, as a result, a simple and valid behavior to group information about the food perception by consumers, including both its impression of sensory descriptors and its acceptance. The technique includes methods such as Open-ended, Check-all-that-apply (CATA), and Emotion Measurement [7].

Although food has an impact on how consumers feel, little is known about how food can be emotionally related to consumers. The search for solutions together with product development context should continue, as it validates studies conclusions related to the subject and allows this tool application and contribution to the food industry [8].

The external preference mapping consists of graphical representations of the descriptive data with those of acceptance by consumers [9]. The correlation of data from external preference mapping is of utmost importance for establishing preference patterns among consumers [10]. In this way, the present work aims to study and characterize samples of conventional diets in sugar guava sweet using Check-all-that-apply and Emotion Measurement methodologies.

Materials and Methods

The research was conducted at Federal Institute of Education, Science and Technology of the South of Minas Gerais–campus Inconfidentes-Minas Gerais State (Brazil). Eight guava sweets samples from different commercial brands, being four conventional (made with sugar) and four samples diets in sugar (made with different sweeteners) were acquired from local markets in Pouso Alegre-Minas Gerais State (Brazil). The conventional guava sweets samples are represented by the letter C, and the diets in sugar ones by the letter D. The sweeteners used in the formulation of the guava sweets were: sorbitol and sucralose (for samples D1, D3 and D4) and maltitol, sorbitol, thaumatin, acesulfame K and sucralose (for sample D2).

Research Ethics Committee

This project was registered in the Brazil platform to obtain a favorable opinion from the research ethics committee in April 2019 and obtained approval in May 2019, with CAAE number 11881019.9.0000.8158. Before performing tests, the consumers presented an identification document with a photo to prove their age of over 18 years and received two copies of the Informed Consent Form (ICF), one copy for the participant and the other for the researcher.

Sensory Analysis

Three sessions of sensory analysis were divulged through printed posters, virtual and oral at IFSULDEMINAS-campus Inconfidentes (Brazil), where all the interested people participated voluntarily, except people who presented some kind of allergy in relation to some of the sample formulation ingredients or who had diabetes mellitus. The conventional diet in sugar guava sweets samples was evaluated by different sensory methods described in the following.

Samples Presentation

The samples were presented in balanced complete blocks [11] in disposable white trays served in napkins coded with three-digit random numbers containing approximately 7 grams of guava sweet with an average size of 3.0 × 2.5 × 1.0 cm in each, at room temperature, in individual booths and illuminated with natural light. It was advised to consume water and wait 30 seconds between each sample consumption to clean the taste buds.

The first and second sessions of the analysis were carried out with 80 consumers, applying the Open-ended method and the Emotions Terms. The third and last session was carried out with 120 consumers to evaluate sensory acceptance attributes, CATA, Ideal CATA, Emotion Measurement, Ideal Test for sweetness and guava flavor, and Purchase Intention. All the analyses were performed with the four conventional and the four diets in the sugar of guava sweet samples.

Descriptors Terms Survey by Open-Ended Method

The analysis counted the participation of 34 men and 46 women between 18 and 52 years old, who described the sample’s sensory characteristics using three to five words or expressions that they judged to be related to each sample [7]. Subsequently, the frequency analysis of responses obtained was performed, and the terms with frequencies higher than 5% were selected for use in the CATA method.

Emotions Terms Survey

Counted the participation of 31 men and 49 women between 18 and 50 years old who described the emotions aroused during the guava sweet samples consumption using three to five words or expressions that they judged to be related to each sample [8].

Sensory Acceptance Test

The test was carried out by 63 women and 57 men between 18 and 64 years old, who evaluated the sample’s attributes of appearance, aroma, flavor, texture, and global impression. In the session, the guava sweet samples were evaluated by a structured nine-point hedonic scale anchored at the ends by “I liked it very much” and “I disliked it very much” [11].

Method Check-all-that-Apply (CATA)

After the frequency analysis of responses obtained by the Open-ended method, the terms with a frequency higher than 5% were used as descriptor terms in the CATA method. The guava sweet samples were supplied to consumers who pointed out in the method sheet all descriptor terms that characterized each sample [12].

Purchase Intention

The guava sweet samples’ purchase intention was also evaluated, using a five-point structured scale ranging from “I certainly wouldn’t buy” to “I certainly would buy” [13].

Just-About-Right Test for Sweetness and Guava Flavor

In order to evaluate the sweetness and guava flavor of the samples, the just-about-right test was carried out using a structured nine-point hedonic scale anchored at the ends “extremely less sweet than ideal” and “extremely more sweet than ideal” for sweetness and “extremely less guava flavor than ideal” and “extremely more guava flavor than ideal” for guava flavor [13]. Pointed out the option that indicated how close to ideal the sweetness and guava flavor was, according to their judgment.

Emotion Measurement

Emotions that were at least 5% frequent in the emotions survey test were used to describe the emotions aroused by consumers when tasting the guava sweet samples. Consumers selected the aroused emotions in themselves during consumption.

Data Analysis

To evaluate the Open-ended method results, descriptor terms with a frequency higher than 5% of responses were used in the CATA method. The same was done in relation to the emotions survey, where emotions with a frequency higher than 5% were used in the Emotion Measurement method. The results obtained in the Acceptance Test and Just-about-right Test were analyzed by ANOVA and Tukey’s tests at 5% probability using the computer software Sensomaker®, developed by [14] Pinheiro, Nunes, and Vietoris, 2013.

A frequency histogram was built for the Purchase Intention Test using Microsoft Office® Excel 2010. Using the global impression results obtained by Acceptance Test and the descriptor terms responses frequency obtained in CATA method, it was possible to build the External Preference Mapping.

Results and Discussion

Open-Ended Method

The sensory evaluation by the Open-ended method, allowed consumers to describe the guava sweet samples using three to five words or expressions. According to the results, it was possible to select 21 terms with a frequency higher than 5% in the answers, presented in Table I.

Table I. Descriptors Terms with a Frequency Higher than 5% at Open-Ended Test for Guava Sweet

No.	Descriptors terms	No.	Descriptors terms	No.	Descriptors terms	No.	Descriptors terms
1	Watery	6	Very sweet	11	Soft texture	16	Sweet taste
2	Bright	7	Slightly sweet	12	Sandness	17	Softness
3	Dark color	8	Weak guava flavor	13	Red color	18	Opaque
4	Hardness	9	Unpleasant flavor	14	Light color	19	Unpleasant flavor
						20	Strong guava flavor
						21	Pleasant texture
5	Bitter taste	10	Sandy texture	15	Creaminess

Emotions Terms Survey

This sensory evaluation allowed consumers to describe three to five emotions aroused during the consumption of guava sweet samples. According to the results, it was possible to select 23 terms with a frequency higher than 5% in the answers, presented in Table II.

Table II. Emotions with Frequency Higher than 5% in the Emotions Terms Survey Test for Guava Sweet

No.	Emotions	No.	Emotions	No.	Emotions	No.	Emotions
1	Affectionate	6	Irritated	11	Astonished	16	Sad
2	Good mood	7	Realized	12	Horrified	17	Lovely
3	In peace	8	Tranquil	13	Unsatisfied	18	Harmonious
4	Grateful	9	Relieved	14	Hurt	19	Happy
						20	Impatient
						21	Angry
5	Bothered	10	Calm	15	Relaxed	22	Nostagilc
						23	Satisfied

Acceptance Test

Table III shows the Acceptance Test results for the attributes of appearance, aroma, flavor, texture, and global impression. For the attribute appearance, the guava sweet samples C1 and C2 did not significantly present a difference (p > 0.05) between them, presenting the highest averages for this attribute. The samples that presented the lowest mean for the attribute appearance were guava sweet D4 and D2 not differing significantly from each other (p > 0.05) but differed from the other samples studied.

Table III. Acceptance Test Attributes of Conventional and Diets in Sugar Guava Sweet Samples

Samples	Appearance	Aroma	Flavor	Texture	Global impression
C1	7.58 ± 1.49 a	7.33 ± 1.53 a	7.36 ± 1.54 a	7.27 ± 1.51 ab	7.46 ± 1.38 a
C2	7.73 ± 1.22 a	7.20 ± 1.52 a	7.15 ± 1.49 a	7.41 ± 1.31 a	7.23 ± 1.55 a
C3	7.47 ± 1.45 ab	7.07 ± 1.52 a	7.10 ± 1.53 ab	7.32 ± 1.50 ab	7.22 ± 1.58 a
C4	6.86 ± 1.79 bc	6.97 ± 1.80 ab	7.24 ± 1.62 a	6.95 ± 1.90 ab	7.07 ± 1.68 a
D1	6.48 ± 1.73 c	6.37 ± 1.80 b	5.53 ± 2.11 c	6.06 ± 2.10 c	5.96 ± 1.94 b
D2	5.46 ± 2.25 d	5.29 ± 2.03 c	4.50 ± 2.34 d	4.55 ± 2.18 d	4.75 ± 2.10 c
D3	7.36 ± 1.31 ab	6.87 ± 1.78 ab	6.41 ± 2.02 b	6.62 ± 1.86 bc	6.77 ± 1.83 a
D4	5.16 ± 2.11 d	4.98 ± 2.03 c	3.60 ± 1.91 e	3.31 ± 2.09 e	3.72 ± 1.80 d

Regarding the attribute aroma, samples C1, C2, and C3 did not significantly present a difference (p > 0.05), obtaining the highest means of acceptance. Samples D2 and D4 showed the lowest accepted with no significant difference (p > 0.05) between them. For the attribute flavor, samples C1, C2, and C4 did not differ from each other (p > 0.05), showing the highest mean values, and sample D4 obtained less acceptability, differing significantly (p ≤ 0.05) from the others. In general, the conventional samples were more accepted for the attribute flavor in relation to the diet samples due to the fact that sucrose enhances fruit flavor in foods since most of the sweeteners used in diet products mask fruit flavor [15].

For the attribute texture sample, C2 obtained the highest hedonic average, not differing significantly (p > 0.05) among the other conventional. The guava sweet sample D4 presented the lowest acceptance, differing significantly from the other samples. According to Alencar et al. [16], the increase in soluble solids content affects the sweet’s texture since it increases the structure hardness.

For global impression, the samples C1, C2, C3, C4, and D3 did not significantly show a difference (p > 0.05) among them and presented the highest means for this attribute. The sample D4 exhibited the lowest acceptability and differed from all others. Mindelo [17] studied the sensory evaluation of guava sweet in mass without sugar using different sweeteners and found that in which the sweeteners substitution (xylitol and sucralose/acesulfame-k) did not affect the product global evaluation comparing to the traditional formulation with sucrose.

Just-About-Right Test for Sweetness and Guava Flavor

Table IV presents the mean values and standard deviations of the just-about-right test for sweetness and guava flavor of conventional diets in sugar guava sweet samples. The average closest to zero is closest to ideality for the attribute being evaluated. According to Table IV all diet guava sweet samples were below the ideal for the attribute sweetness, with opposite result for all conventional guava sweet samples. For the guava flavor attribute all samples were below the ideality.

Table IV. The Mean Values and Standard Deviations for Sweetness and Guava Flavor of Conventional Diets in Sugar Guava Sweet Samples

Samples	Sweetness	Guava flavor
C1	0.23 ± 1.15 ab	−0.21 ± 1.23 a
C2	0.71 ± 1.48 a	−0.03 ± 1.46 a
C3	0.42 ± 1.47 a	−0.24 ± 1.76 a
C4	0.43 ± 1.27 a	−0.39 ± 1.06 a
D1	−1.23 ± 1.29 c	−1.37 ± 1.42 b
D2	−1.44 ± 1.66 c	−1.69 ± 1.54 b
D3	−0.17 ± 1.61 b	−0.39 ± 1.55 a
D4	−2.60 ± 1.32 d	−2.33 ± 1.57 c

The D3 sample was closer to ideal for sweetness, differing from all except the conventional C1, which was closer to ideality. Sample D3 is the only one that presents lactic acid as an acidulant, and for this reason, it has a lower acid taste compared to the others and, thus, better sweetness perception. In the others, citric acid is used more for its pleasant taste and immediate perception. According to Almeida et al. [18], sweets made with sweeteners show a lower acceptance tendency compared to high-caloric value products.

For guava flavor attribute sample C2 proved to be closer to the ideal, not differing significantly from C1, C3, C4, and D3, as it is the only conventional sample that was made with guava pulp. Sample D4 showed guava flavor more distant than ideal, differing significantly (p < 0.05) of all of them, being the only one that does not use acidulant in its formulation, and consequently less highlighting the natural fruit aroma [19]. Moreover, food is directly influenced by its aroma and so D4 has less intensity of guava flavor. Siqueira [20] evaluated the physical-chemical and sensory characterization of light in sugar guava sweet in mass and verified that higher guava flavor perception, higher guava sweet acidity perception.

Purchase Intention

As described by Meilgaard et al. [13], the purchase intention test shows the consumers’ attitudes towards buying a product. Thus, Fig. 1 shows the frequency distribution of the purchase intention for all samples.

Fig. 1. Frequency distribution for purchase intention of conventional and diets guava sweet samples. Source: The author, 2019.

The purchase intention of samples is presented in Fig. 1. The guava sweet samples C1 and C4 presented higher frequencies of positive purchase intention, represented by the sum of the answers “I certainly would buy” and “I probably would buy” with 69% and 60%, respectively. The purchase indecision, represented by the answer “maybe I would buy,” had higher frequency for D3 and C2, with 34% and 33%, respectively. Regarding the negative purchase intention, represented by the sum of the answers “I certainly wouldn’t buy” and “I probably wouldn’t buy,” D4 and D2 obtained the highest frequencies with 85% and 65%, respectively. The results corroborate with the acceptance test and ideal for sweetness and guava flavor, where sample C1 was the most accepted for aroma, flavor, and global impression, and sample D4 was the least accepted for all attributes, besides being more distant from ideal for both issues.

Mindelo [17] studied the microbiological and sensory evaluation of guava sweet in paste without sugar using different sweeteners and verified that the purchase intention test showed positive results, particularly for diet formulations.

External Preference Mapping for Check-all-that-Apply (CATA)

Fig. 2 presents the results of the CATA test and their correlation with the overall impression for all samples.

Fig. 2. External preference mapping with global impression results, CATA, and ideal CATA for conventional and diets in sugar guava sweet samples. Source: The author, 2019.Note: WAT = watery; BRI = bright; DCO = dark color; HAR = hardness; BIT = bitter taste; VSW = very sweet; SSW = slightly sweet; WGF = weak guava flavor; UPF = unpleasant flavor; SAT = sandy texture; SOT = soft texture; SAN = sandness; REC = red color; LIC = light color; CRE = creaminess; SWT = sweet taste; SOF = softness; OPA = opaque; PLF = pleasant flavor; SGF = strong guava flavor; PLT = pleasant texture; IDCON = conventional ideal; IDDIET = diet ideal. ** PC1 + PC2 represent 80.57% of the variability between samples.

The results of CATA and ideal CATA tests and the correlation with global impression results are shown in Fig. 2. The first and second dimensions of mapping correspondence analysis were responsible for 80.57% of experimental data variance, being 61.60% and 18.97%, respectively. The total variation in the two principal components was over 70%, as recommended by Muñoz et al. [21].

The External Preference Mapping derived from the global impression and CATA test allows the visualization at sample acceptance level based on the descriptor terms. Thus, it is noted that sample D2 was characterized by the term “watery,” corroborating with the lower acceptability of attribute texture and with negative purchase intention frequency.

Sample C4 is close to the term “light color”, corroborating with positive purchase intention frequency and good acceptability for attribute appearance in the acceptance test, making it evident that part of consumers have a preference for lighter coloration guava sweet. Sample D4 is characterized by the term “hardness,” validating with the lowest acceptance for all attributes and with the highest negative purchase intention. The other samples were not close enough to the terms to be characterized. For the global impression results related to the ideal CATA test, C2 was closer to ideal for conventional guava sweet, matching the highest acceptance for appearance and texture. Consumers consider attributes such as appearance and texture extremely important to characterize conventional guava sweets as ideal. The sample D3 was closer to ideal for a diet in sugar guava sweet, supporting the acceptance test.

A similar study was conducted by Peçanha et al. [22], who presented the external preference mapping and identified the sensory characteristics of guava cascade sweet as the attributes: red color, guava cascade sweet aroma, guava sweet flavor, among others. It was verified the preference for products with less intensity and firmness in chewing resistance.

External Preference Mapping for Emotion Measurement

Fig. 3 presents the results of the Emotional Response test and their correlation with the overall impression for all samples.

Fig. 3. External preference mapping with the results of global impression and emotion measurement for conventional and diets in sugar guava sweet samples. Source: The author, 2019.Note: AFE = affectionate; GOM = good mood; IPE = in peace; GRA = grateful; BOT = bothered; IRR = irritated; REA = realized; TRA = tranquil; REL = relieved; CAL = calm; AST = astonished; HOR = horrified; UNS = unsatisfied; HUR = hurt; REL = relaxed; TRI = sad; LOV = lovely; HAR = harmonious; HAP = happy; IMP = impatient; ANG = angry; NOS = nostalgic; SAT = satisfied. ** PC1 + PC2 represent 89.91% of the variability between samples.

Fig. 3 presents the results of the emotion measurement test and its correlation with global impression results for conventional diets in sugar guava sweet samples. The first and second dimensions of external preference mapping correspondence analysis were responsible for 89.91% of experimental data variance, 81.76%, and 8.15%, respectively. The total variation in the two principal components was higher than 70%, as recommended by Muñoz et al. [21].

According to Fig. 3, sample C1 was represented by the emotion “lovely,” C2 by the emotion “nostalgic,” and C4 by the emotion “relieved,” showing proximity to positive emotions and corroborating with the acceptance test (C1 and C2 exhibited the highest hedonic averages for appearance, aroma, flavor, texture and global impression); purchase intention test (C1 and C4 obtained the highest positive frequencies); ideal test (C1 was closer to ideal for sweetness and C2 was closer to ideal for guava flavor) and external preference mapping for CATA and ideal CATA tests, due to the correlation of the term “light color” with C4, and C2 being closer to ideal for a conventional guava sweet.

Sample D2 is represented by the emotions “astonished” and “unsatisfied,” and D4 was close to the negative emotions: horrified, impatient, angry, irritated, unconfirmed, hurt, and sad. These results are related to the acceptance test (showed the lowest hedonic averages for appearance, aroma, flavor, texture, and global impression), purchase intention test (showed the highest negative frequencies), ideal test for sweetness and guava flavor (these samples were more distant from the ideal) and regarding the external preference mapping to CATA test, due to the relation with the terms “watery” and “hardness,” respectively. The other samples did not present considerable proximity to the terms to be characterized.

Chaya et al. [23] studied the emotion measurement of 10 lager-type beers found in the United Kingdom. They reported 17 emotions, such as active, bored, enthusiastic, free, friendly, and happy, among others, being able to differentiate commercial beers and provide a rich vision of consumer perception. Melo et al. [24] evaluated commercial samples of conventional and organic olive oil by the emotional responses method. They observed that organic samples were represented by the descriptors well, indifferent, disinterested, unsatisfied, and disgusted.

Conclusion

According to the external preference mapping for the Check-all-that-apply test and Emotion Measurement, sample C4 showed higher proximity with one positive emotion descriptor term and, C1 and C2 were close to positive emotions. In contrast, D2 and D4 were close to negative descriptor terms and emotions. These results are in accordance with acceptance tests, purchase intention, and just-about-right tests.