Sensory Characterization and Acceptance of Amazonian Robustas Coffee Brews by Consumers Using a Home-Use Test

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This study explored sensory characteristics and acceptance of two Coffea canephora intervarietal hybrids, as natural and fermented versions. A home-use test of acceptance and the CATA method were applied to the four samples. A total of 127 participants brewed the coffee samples and evaluated the brews at their homes. The sample size is adequeate and allows segmentation of the participants, which the authors have done. Some chemical analyses were also done for the coffees, but it is not clear how they are related to the aims of the study or the sensory analysis.

The main result was that the naturally processed coffees were liked more that the fermented ones. My major concern is that the conclusions related to the sensory characteristics are not fully supported by the results of the CATA profiling.


Specific issues:

1. Abstract. It could be useful for a reader to add the number of participants. Please also think whther you could say that natural coffees were "preferred by almost half of the participants, mainly women". You did not run a preference test (ranking, which is best).

2. Methods, line 110. In the supplementary table you cite a personal communication for the sensory descriptors were taken (Alves 2022). Consider citing the communication also here in the body text, to make it clear for a reader that you did not study it in the present study. Consider also describe a little more about how the expert panel defined the sensory attributes/notes you listed in the Table S1 (e.g., how many expert panellists were included).

3. Methods, 2.2. Here you describe a chromatographic method. However, you do not tell here (nor in the Introduction (aims, lines 95-98)), which components you will analyze, until lines 160-161. Consider telling the compounds (or their categories) earlier in the text. I think listing standards (lines 147-148) implies, but does not make it sure that you determined these compounds. Also, consider adding the rationale and aims for the chemical analysis.

4. Methods, lines 230-232. It seems that you used a line scale ("continuous"). If so, please make it clear. With numbers it could be confused with a category scale (with limited number of response options).

5. Methods, 2.3.2. When the descritors for the CATA task were taken from the previous studies (cited appropriately), what is the meaning of the information on senroy attributes included in the Table S1 (when they are not used for the CATA)? Consider whether Table S1 is needed, especially when it is based on personal communication that cannot be tracked to a specific published study.

6. Methods, 2.3.2. I assume that the original questionnaire for the sensory analysis was not in English. Please report the language used. In CATA, the attribute pairs "bitter" and "slightly bitter" etc. are problematic. Does "slightly bitter" indicate absence of bitterness or than the coffee was a little bitter? Could a participant have checked both "slightly bitter" and "bitter", if the coffee was clearly bitter? Please explain.  

7. Figure 1. Which analysis this plot is based on?

8. Conclusions. Please re-think the conclusions, which conclusions the data really support. Can you really say that beverages were characterized by "slightly sweet" taste, when less than a quarter of the participants selected this attribute to apply for a sample (max. citation frequency was 0.228 in the case of BRS 2357N)? Attribute "caramel" is even more problematic. When less than 15% of the participants thought it applied to a sample, could you say "Natural coffee brews -- were mostly associated with caramel flavor --"). It could be the case that the natural samples were more often that the other samples characterized as sweet and caramel, but it still does not mean that most of the participants think that the the natural samples have sweet and caramel notes.

Comments on the Quality of English Language

As far as I can tell, this manuscript is written in fluent English.

Author Response

ID: beverages-3060442: Sensory characterization and acceptance of Amazonian Robustas coffee brews by consumers using a home-use test

Authors: Thayna Viencz, Claudimara da Silva Portela, Rodrigo Barros Rocha, Enrique Anastácio Alves, André Rostand Ramalho, Rafael Carlos Eloy Dias, Marta de Toledo Benassi

Section: Tea, Coffee, Water, and Other Non-Alcoholic Beverages

 

We want to thank the editor and reviewers. The suggestions were accepted and included in the new version of the manuscript. The revisions were highlighted in yellow so the editor and reviewers could quickly check the modifications. The changes are described below.

 

Editor: Regarding overlap with other articles

During our standard checks of all submissions, we noticed that the following sections had significant overlap with previously published articles, and we would therefore request you to rephrase these paragraphs during your revision stage. **Section 2.2/Section 2.2.1**. Please revise your manuscript according to the iThenticate report attached and make sure that there is no large part repetition with the published paper.

Answer: We rephase the paragraps as suggested (pages 3 and 4, lines 139 to 188).

 

Comments on the Quality of English Language

Reviewer 1: As far as I can tell, this manuscript is written in fluent English.

Reviewer 2: The quality of the English language in the manuscript is commendable. The text is clear, precise, and well-structured, ensuring readability and comprehension. The scientific terminology is used appropriately, demonstrating a strong command of the subject matter.

Answer: Thank you for the compliments. Considering this, only minor changes were made to the English language to adapt the text to the reviewers' suggestions.

 

Reviewer 1

Comments and Suggestions for Authors

This study explored sensory characteristics and acceptance of two Coffea canephora intervarietal hybrids, as natural and fermented versions. A home-use test of acceptance and the CATA method were applied to the four samples. A total of 127 participants brewed the coffee samples and evaluated the brews at their homes. The sample size is adequate and allows segmentation of the participants, which the authors have done.

Some chemical analyses were also done for the coffees, but it is not clear how they are related to the aims of the study or the sensory analysis.

The main result was that the naturally processed coffees were liked more that the fermented ones. My major concern is that the conclusions related to the sensory characteristics are not fully supported by the results of the CATA profiling.

Answer: Thank you for the insightful comments; compliance with these suggestions will be discussed in each point below.

 

Specific issues:

1. Abstract. It could be useful for a reader to add the number of participants. Please also think whether you could say that natural coffees were "preferred by almost half of the participants, mainly women". You did not run a preference test (ranking, which is best).

Answer: The information on the number of participants was included (page 1, line 14)

We understand the reviewer's concern with the use of "preference". In fact, we ran an acceptance test and inferred the preference for natural coffees, after segmentation, for this group of participants based on the multivariate analysis since they were located near these samples in the plot. We believe it was clear in the discussion (page 8), but we agree with the reviewer that it was less clear in the abstract and conclusion, so we revised the writing, pointing out that the samples were accepted (page 1, lines 19, 22 and 23; page 12, lines 458, 462 and 464).

 

2. Methods, line 110. In the supplementary table you cite a personal communication for the sensory descriptors were taken (Alves 2022). Consider citing the communication also here in the body text, to make it clear for a reader that you did not study it in the present study. Consider also describe a little more about how the expert panel defined the sensory attributes/notes you listed int he Table S1 (e.g., how many expert panelists were included).
3. Methods, 2.3.2. When the descriptors for the CATA task were taken from the previous studies (cited appropriately), what is the meaning of the information on sensory attributes included in the Table S1 (when they are not used for the CATA)? Consider whether Table S1 is needed, especially when it is based on personal communication that cannot be tracked to a specific published study.

Answer: We answered these two questions together since the subject was related. We agree that our motivation to present the data and the presentation format were not well described.

We included the citation in the text to clarify that we are using the information, but it was not obtained in this research (page 3, line 112).

The reviewer is correct; the expert's analysis was lacking in information. We included information on the protocol used in the text and regarding the data itself in the Table S1 footnote.

Regarding the suggestion to consider whether Table S1 is necessary, we prefer to keep it even if it is based on personal communication; the data are not published yet but come from a reliable source (they are part of an extensive study conducted by the coauthor, Enrique Alves). We do not use all these attributes on CATA directly because experts' terms could be challenging for consumers, but they were also useful in helping us define our list.

To justify our option to keep the table:.

- First, following the reviewer's suggestions, we believe the information is now more precise and more useful.

- Second, as we pointed out to reviewer 2, one goal of our research is to describe the effect of fermentation on consumers' acceptance when preparing their C. canephora coffees at home. For Arabica coffees, the literature described that fermentation impacts both sensory characteristics and acceptance, so this process has been used to "improve" the cup quality. In previous research of our group (with the same samples applying descriptive sensory tests on a lab scale, in Viencz et al., 2024), we show that the fermentation process changes the sensory characteristics of the beverages, regardless of the genetic difference. Based on the information in Table 1S, we can see that, in experts' opinions, fermentation also led to an increase in quality scores for these Amazonian Robusta coffees. Nevertheless, looking at our results, we see that fermentation actually impacts the sensory characteristics perceived by consumers, but the process would not necessarily improve the general acceptance. However, we can also observe that some groups of consumers appreciate the characteristics of fermented C. canephora coffees, so there may be a niche for differentiated products.

- Third, it is interesting that considering natural coffees, clone BRS 2314 — which received higher quality scores from experts (Table S1) — is equally accepted by consumers as BRS 2357, reinforcing the need for consumer studies to complement expert quality assessments.

So, we believe that Table S1 has valuable information. Comparing our results with experts' opinions highlights the importance of knowing the consumer's point of view.

 

3. Methods, 2.2. Here you describe a chromatographic method. However, you do not tell here (nor in the Introduction (aims, lines 95-98)), which components you will analyze, until lines 160-161. Consider telling the compounds (or their categories) earlier in the text. I think listing standards (lines 147-148) implies but does not make it sure that you determined these compounds. Also, consider adding the rationale and aims for the chemical analysis.

Answer: It is an excellent suggestion; information on the compounds analyzed and the aim of studying them were included in the introduction (page 3, lines 98-101).

We understand the reviewer's concern regarding the validity of the chromatographic methods applied, but we are very confident about the reliability of the analysis. Our research group (two of the coauthors of this study, Rafael Dias and Marta Benassi) developed and validated both methodologies in previous articles (10.1590/S0100-40422006000600003, 41 citations in Scopus; 10.1021/jf9027427, 70 citations in Scopus). The reference (Viencz et al., 2023) cited as containing the description of both analyses adapted to the studied matrix (roasted C. canephora coffees) is also coauthored by some of the authors of this study and is published in a journal that is mainly concerned with the quality of analytical methods and data obtained (Journal of Food Composition and Analyses).

 

4. Methods, lines 230-232. It seems that you used a line scale("continuous"). If so, please make it clear. With numbers it could be confused with a category scale (with limited number of response options).

Answer: The reviewer is correct. The term hybrid indicates that it is a linear scale with anchors of verbal affective labels, but it could be unclear to those unfamiliar with the concept. The information was included (page 5, lines 237-238); as the original reference that proposed the scale is available, we believe that the reader can check in case of any additional doubts.

 

6. Methods, 2.3.2. I assume that the original questionnaire for the sensory analysis was not in English. Please report the language used. In CATA, the attribute pairs "bitter" and "slightly bitter" etc. are problematic. Does "slightly bitter" indicate absence of bitterness or than the coffee was a little bitter? Could a participant have checked both "slightly bitter" and "bitter", if the coffee was clearly bitter? Please explain.

Answer: It was in Portuguese; we included the information (page 5, lines 244-245).

Regarding the use of the term slightly bitter on CATA, it indicates that the coffee was a little bitter. The bitter, sour, and sweet tastes are always associated with the consumption of coffee beverages. However, they are presented/perceived in different levels of intensity that are determined by the composition of the beverage, which depends on the genetics (species and cultivars), and post-harvest, roasting, and brewing processes. If we just use the attributes bitter, sour, and sweet in CATA, there is a chance that a slightly bitter sample could not be associated with bitter since the consumer could choose other attributes that he/she thinks are more dominant. When the option of slightly bitter is offered, he/she could consider catching the term. We can not prevent one assessor from indicating both alternatives (bitter/slightly bitter), but this did not occur in our research; we believe that there is an understanding of using these intensity terms because it is quite common in CATA analysis.

As an example, a well-recognized paper (70 citations in Scopus) by Prof. Gaston Ares's research group (Bruzone et al. 2015; 10.1016/j.foodqual.2015.04.017) described the use of intensity modifiers […The terms included were selected… and their opposites, generated by preceding an intensity adjective (... Not very and Not much): thick, not very thick, …, smooth, not very smooth, sweet, not very sweet, vanilla flavor, not much vanilla flavor…]. These authors also reported that using opposite terms can help check the validity of consumers’ responses to the CATA question.

The same can be observed in other articles: sweet/barely sweet (Ares et al. 2017; 10.1016/j.foodqual.2016.08.011), vanilla flavor/not much vanilla flavor, creamy, not very creamy, sweet/not very sweet (Ares et al. 2015, 10.1016/j.foodqual.2015.05.007). For the specific use for coffee products, we can cite two examples: Mori et al. 2018 (10.1590/S0100-204X2018000900010) used bitter, slightly bitter, sweet, slightly sweet; Dauber et al. 2024 (10.1016/j.afres.2023.100373) used very low sweetness, too sweet.

 

7. Figure 1. Which analysis this plot is based on?

Answer: The reviewer is correct; the figure must be self-explanatory. The information on the statistical methods used was included (page 8, lines 342-343).

 

8. Conclusions. Please re-think the conclusions, which conclusions the data really support. Can you really say that beverages were characterized by "slightly sweet" taste, when less than a quarter of the participants selected this attribute to apply for a sample (max. citation frequency was 0.228 in the case of BRS 2357N)? Attribute "caramel" is even more problematic. When less than 15% of the participants thought it applied to a sample, could you say "Natural coffee brews – were mostly associated with caramel flavor --"). It could be the case that the natural samples were more often that the other samples characterized as sweet and caramel, but it still does not mean that most of the participants think that the the natural samples have sweet and caramel notes.

Answer: We agree with the reviewer. There is no need to conclude on less characteristic things. We changed what was reported in the conclusion and in other points of the manuscript.

The description of "slightly sweet" for the beverages in general was removed from the abstract, main text, and conclusion (page 1, line 17; page 10, line 386; page 12, line 457). The same was done for the attributes "caramel and sweet" regarding natural coffees; the information was removed from the abstract and conclusion (page 1, line 19; page 12, line 459) and rewrote in the main text (page 11, lines 417-418) as suggested by the reviewer.

Reviewer 2 Report

Comments and Suggestions for Authors

The study on sensory characterization and acceptance of Amazonian Robustas coffee brews demonstrates several strengths and weaknesses. Notably, it addresses a highly relevant topic within the coffee industry by examining consumer acceptance of Coffea canephora hybrids, reflecting the growing interest in specialty coffees and sustainable production. The methodological choice of using a home-use test (HUT) enhances the ecological validity by replicating real-world consumption conditions, and the comprehensive analysis employing both hedonic scaling and Check-All-That-Apply (CATA) methodologies offers detailed insights into consumer preferences. However, the study's sample size of 127 participants, primarily with a high level of education, may not fully represent the general coffee-drinking population, thus limiting generalizability. While the statistical methods like ANOVA and Cochran’s Q test are appropriate, the study could benefit from advanced multivariate analyses to explore interactions between sensory attributes and demographics. Additionally, potential biases in home-use tests due to variability in preparation methods could be better controlled. The study is logically structured, with clear sections and appropriate scientific language, ensuring clarity and coherence in presenting the research problem and findings. I have a few questions as follows:

1. Why did you choose to focus specifically on the clones BRS 2314 and BRS 2357 for this study?

2. Why was the home-use test (HUT) deemed more appropriate than laboratory-based sensory evaluations for this study?

3. Why might there be differences in sensory acceptance between natural and fermented versions of the same coffee clones?

4. Why do you think the educational level of participants influenced their preferences for different coffee brews?

5. Why were certain sensory attributes, such as fruity and tobacco flavors, more associated with fermented coffees compared to natural ones?

Comments on the Quality of English Language

The quality of the English language in the manuscript is commendable. The text is clear, precise, and well-structured, ensuring readability and comprehension. The scientific terminology is used appropriately, demonstrating a strong command of the subject matter.

Author Response

ID: beverages-3060442: Sensory characterization and acceptance of Amazonian Robustas coffee brews by consumers using a home-use test

Authors: Thayna Viencz, Claudimara da Silva Portela, Rodrigo Barros Rocha, Enrique Anastácio Alves, André Rostand Ramalho, Rafael Carlos Eloy Dias, Marta de Toledo Benassi

Section: Tea, Coffee, Water, and Other Non-Alcoholic Beverages

 

We want to thank the editor and reviewers. The suggestions were accepted and included in the new version of the manuscript. The revisions were highlighted in yellow so the editor and reviewers could quickly check the modifications. The changes are described below.

 

Comments on the Quality of English Language

Reviewer 1: As far as I can tell, this manuscript is written in fluent English.

Reviewer 2: The quality of the English language in the manuscript is commendable. The text is clear, precise, and well-structured, ensuring readability and comprehension. The scientific terminology is used appropriately, demonstrating a strong command of the subject matter.

Answer: Thank you for the compliments. Considering this, only minor changes were made to the English language to adapt the text to the reviewers' suggestions.

 

Reviewer 2

 

Comments and Suggestions for Authors

The study on sensory characterization and acceptance of Amazonian Robustas coffee brews demonstrates several strengths and weaknesses. Notably, it addresses a highly relevant topic within the coffee industry by examining consumer acceptance of Coffea canephora hybrids, reflecting the growing interest in specialty coffees and sustainable production. The methodological choice of using a home-use test (HUT) enhances the ecological validity by replicating real-world consumption conditions, and the comprehensive analysis employing both hedonic scaling and Check-All-That-Apply (CATA) methodologies offers detailed insights into consumer preferences. However, the study's sample size of 127 participants, primarily with a high level of education, may not fully represent the general coffee-drinking population, thus limiting generalizability. While the statistical methods like ANOVA and Cochran's Q test are appropriate, the study could benefit from advanced multivariate analyses to explore interactions between sensory attributes and demographics. Additionally, potential biases in home-use tests due to variability in preparation methods could be better controlled. The study is logically structured, with clear sections and appropriate scientific language, ensuring clarity and coherence in presenting the research problem and findings. I have a few questions as follows:

Answer: Thank you for your comments and praise for our work. We will discuss compliance with these suggestions for each point below.

 

1. Why did you choose to focus specifically on the clones BRS2314 and BRS 2357 for this study?

Answer: This option was made as conducting the previous study described in the Introduction (Viencz et al., 2024). We selected the two most differentiated clones of the cultivar (considering the phenotypic and genotypic characteristics) and also included the fermentation variable. As we observed that the assessors discriminated both the genetics and the presence of the fermentation process and described specific attributes for each beverage, we chose to carry out the logical sequence in this research: to verify whether this could also be observed by consumers when preparing their coffees at home, and if they were accepted. The sentence in the Introduction was rephrased to make it clear (page 2, lines 87-89).

 

2. Why was the home-use test (HUT) deemed more appropriate than laboratory-based sensory evaluations for this study?

We saw that there is also a comment associated with this matter in the reviewer's general comments (Additionally, potential biases in home-use tests due to variability in preparation methods could be better controlled), so we are answering these two questions together.

Answer: We consider the HUT the proper choice for the study.

We understand the reviewer's concern about potential biases in home-use tests due to variability in the brewing. We believe that is an onus and a bonus of this kind of study. We did our best to ensure the standardization of the process, providing the kit with detailed instructions, including a test sample to be used in the first preparation to familiarize themselves with the brewing method. The participants also received a video carrying out preparation to help clarify doubts about the coffee maker use (this information was included in the methodology section, page 5, lines 232-234). One of the goals of the home-use test is to allow participants to actively prepare and consume the product as part of their daily routine, and it includes little variations in the preparation (compared to a lab-controlled experiment). This could be compared to the robustness of an analytical method, which is not affected by minor variations in conditions. Therefore, if our final results were coherent and segmentation was observed between groups of consumers, we believe that the procedure was robust and the standardization was sufficient.

Another important point we forgot to mention (now included in the methodology section, page 5, lines 239-241) is that the assessors could also consider perceptions obtained during the preparation, which is particularly relevant for an aromatic beverage such as coffee. The aroma during preparation is a key point during home preparation regarding the impression of the beverage that could never be observed in a lab-scale test.

As additional information, the research was carried out during the COVID-19 pandemic, so carrying out the test at home also allowed us to work with the social distancing necessary at that time to ensure the security of the assessors. These procedures were described to obtain permission from the Human Research Ethics Committee to conduct the study. This information was also included in the text (page 4, lines 193-195).

Recently, an interesting HUT case study for brewed coffee was published (Park et al. 2024; 10.1016/j.foodres.2024.114349; that was included in the manuscript) comparing conventional face-to-face central location test (CLT) conditions with two non-face-to-face HUT test conditions. One home-use test (C-HUT) provided consumers with sensory evaluation protocols in a box with no direct interaction (more similar to our study) and the other (O-HUT) additionally asked participants to participate in an online meeting to prepare samples and conduct sensory evaluations by following a live guideline by the researcher. The sensory attributes used to significantly differentiate the samples were similar for all three conditions. So, the authors concluded that both C-HUT and O-HUT are suitable methods for collecting sensory data and overcoming face-to-face contact limitations.

 

3. Why might there be differences in sensory acceptance between natural and fermented versions of the same coffee clones?

Answer: As described in question 1, in a previous study (Viencz et al., 2024) with descriptive sensory tests in a lab scale, we observed that the assessors discriminated both the genetics and the presence of the fermentation process and described specific characteristics for each beverage. Of course, differences in characteristics do not necessarily imply differences in acceptance, which is why this study was conducted. For Arabica coffees, there is much information in the literature that the fermentation process impacts both the sensory characteristics and acceptance; fermentation has been used to "improve" the cup quality. The same behavior was observed for C. canephora coffees, considering expert opinions (see Table 1S for an increase in quality scores); so, we hypothesized that the same could happen when consumers evaluate the coffee brews. We found that, for this species, the fermentation actually impacts the sensory characteristics perceived by consumers, but the process would not necessarily improve the general acceptance. It probably occurs because the consumer is not used to the regular C. canephora beverage characteristics, and we added yet another modification to the matrix (with fermentation), which may have caused some strangeness. However, we can also observe that some groups of consumers appreciate the characteristics of fermented C. canephora coffees, so there may be a niche for differentiated products. It is interesting information for producers and coffee chains to consider.

 

4. Why do you think the educational level of participants influenced their preferences for different coffee brews?

We saw that there is also a comment associated with this matter in the reviewer's general comments (However, the study's sample size of 127 participants, primarily with a high level of education, may not fully represent the general coffee-drinking population, thus limiting generalizability), so we are answering these two questions together.

 

Answer: The reviewer is correct. The high level of education of the participants is a limitation of the study, so we included an observation in the discussion section (page 12, lines 451-453) (Regarding the study's limitations, it should be considered that our research was limited to participants mostly aged between 20 and 60 and with a high level of education; future studies could address a broader variation in the consumer profile).

Even with this limitation, we think that it was interesting to observe some possible indications of preference niches for fermented coffees. As we described, the group who preferred the BRS 2314F, was composed of younger participants with a higher level of education; the group who preferred the BRS 2357F was composed of older participants with a low degree of education and a higher coffee consumption frequency. However, considering the sample size and the limitations we described, we prefer not to make broader statements on this point, and waiting for future studies could address a wider variation in the consumer profile.

 

5. Why were certain sensory attributes, such as fruity and tobacco flavors, more associated with fermented coffees compared to natural ones?

Answer: We agree that this point was not properly discussed, so we included information (pages 11 and 12, lines 420-427, 431-437) with several references.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

I thank the authors for carefully considering my comments, revising the manuscript accordingly, and providing appropriate responses. I regard the manuscript as much improved and have no further comments.

Comments on the Quality of English Language

English language is fine.

Author Response

We are grateful for the reviewers' positive feedback regarding our response to the suggestions for improving our manuscript. Indeed, the manuscript is now in a much-improved state.

Reviewer 2 Report

Comments and Suggestions for Authors

The authors have effectively addressed the comments and suggestions, significantly enhancing the manuscript's quality. Their responses demonstrate a clear understanding of the methodological choices, such as the use of HUT for its ecological validity and the robust standardization procedures to minimize biases. They have provided thorough explanations for the selection of specific coffee clones and the implications of fermentation on sensory characteristics and consumer acceptance. Additionally, the authors acknowledged the study's limitations, such as the educational level of participants, and proposed areas for future research to increase generalizability. Overall, their revisions reflect a thoughtful and scientific approach, improving the study's clarity and rigor.

Comments on the Quality of English Language

The quality of the English language is well-written and uses appropriate scientific terminology, enhancing the overall readability and professionalism of the work.

Author Response

We are grateful for the reviewers' positive feedback regarding our response to the suggestions for improving our manuscript. Indeed, the manuscript is now in a much-improved state.