Panel performance, sensory characterization and consumer preference of Chinese traditional chrysanthemum drink

doi:10.21203/rs.3.rs-3087151/v1

Download PDF

Research Article

Panel performance, sensory characterization and consumer preference of Chinese traditional chrysanthemum drink

https://doi.org/10.21203/rs.3.rs-3087151/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

In China, herbal drinks possess long-standing traditional cultural characteristics cater to consumers' demand for both natural ingredients and functional benefits. So far, the sensory properties of herbal beverages have not been subjected to descriptive analysis. Here, we selected 12 high-selling herbal beverages containing chrysanthemum samples on China's largest online shopping website. A total of 11 sensory descriptors have been associated with the samples, while assessed panel performance and sensory characteristics of samples by PanelCheck software. In addition, using ConsumerCheck to investigate the consumer acceptance of the same samples, applying preference mapping analysis to address the relationship between descriptive and consumer liking date. Box plot and Stacked histogram visualise the distributions of the liking ratings across all consumers for each of the tested products. Preference mapping revealed that the main sensory attributes driving consumers' preferences are fragrant sweet flavour, brillancy, lubrication, sweet taste and overly sweet taste. Thus, this study could guide the development of Chinese traditional chrysanthemum drink. At the same time, the results also provided a simple and open-source software to data statistical method for practitioner without commercial software and programming skills.

herbal beverages

chrysanthemum

descriptive sensory analysis

acceptability

PanelCheck

ConsumerCheck

Chrysanthemum, originating in China and cultivated for over 2,500 years, is one of the top ten renowned flowers in China. Chrysanthemum is not only a rare ornamental flower, but also a homologous plant of medical and edible food (Luo et al. 2017). In China, chrysanthemums are harvested in batches during the period of September to November when they are in full bloom. The fresh chrysanthemums undergo a drying process to become dry chrysanthemum, which can be drinking after brewing for consumers (Yuan et al. 2015; Bae et al. 2008). Chrysanthemum-based beverages are highly favored among Chinese consumers. As early as 1596, Li Shizhen, a medical expert of the Ming Dynasty, recorded the efficacy of chrysanthemum in clearing heat and toxic materials in his“Compendium of Materia Medica”. In 1694, Wang Ang's “Notes on Materia Medica” documented that chrysanthemum possessed a dual flavor profile of bitterness and sweetness and beneficial for lung and kidney health. Modern pharmacological research has demonstrated that chrysanthemum possesses diverse various bioactivities, including anti-inflammatory, antioxidant, antitumorigenic, hepatoprotective and cardiovascular protective effects (Liao et al. 2023; Li et al. 2023; Zhan et al. 2022).

In the past decade, with the continuous enhancement of health consciousness among Chinese consumers, the concept of healthy diet in the food and beverage industry has gradually taken root in people's minds. Natural and healthy beverages have garnered increasing attention from consumers. In China, herbal drinks possess long-standing traditional cultural characteristics, and the health functions of “medicine and food homologous” herbs have been recognized in the consumer market (Deng et al. 2021). Herbal beverages have become pivotal in the development of the beverage industry as they cater to consumers' demand for both natural ingredients and functional benefits. Currently, numerous renowned food companies in China, including CocaCola, Vita Food, Dinjin Food and Healing Forest, have introduced herbal beverages infused with chrysanthemum to their product lines. Based on this, the present study collected high-selling herbal beverages containing chrysanthemum from the Chinese online market, and applied sensory methods to investigate their sensory characteristics and consumer preferences. The findings can provide guidance for developing market-oriented chrysanthemum beverages.

In sensory evaluation research activities, descriptive sensory analysis and consumer acceptance testing are typically conducted separately. Panel of 8 to 12 trained assessors usually use descriptive methods to describe products of interest (A González et al. 2021). This methodology provides very accurate and reproducible results (Faccioli et al. 2021). Consumer acceptance testing are only the consumers of the tested sample category and may reflect the degree of acceptance of a particular product by general consumers (Bhardwaj et al. 2022). Generally, more than 50 evaluators are selected (Tzompa-Sosa et al. 2022). Recently, many researchers have begun to connect consumer acceptance testing with descriptive sensory analysis in order to provide better guidance for sample development and formulation improvement. Mauricio A A, et al. (2016) studied correlations between sensory properties and acceptance about carrot cakes, obtained consumer acceptance drivers of a gluten-free baked food. Ubeda C, et al. (2017) obtained external preference mapping by PLS analysis, revealed that the main sensory attributes driving consumers' preferences are raisin, toasted caramel, spicy, and liqueur aroma. Lattey K A, et al. (2010) investigated sensory attributes and consumer acceptance that determined what sensory attributes most drive consumer and expert acceptance for Cabernet Sauvignon and Shiraz wines. At present, the main data statistical methods are mainly simple descriptive statistics, principal component analysis (PCA), partial least squares regression (PLSR), principal component regression (PCR) and conjoint analysis (Jo et al. 2020; Campoceballos et al. 2023; Mudasser et al. 2021; Huang et al. 2010; Martins et al. 2017). These statistical methods have been long available, either through commercial software, or as part of free statistical packages in open source programming languages such as Python or R (Ho 2015; Jérme et al. 2013). These methods present some difficulties for sensory analysts without commercial software and programming skills. The PanelCheck software, which was released in 2007, is an open source software that can be downloaded, distributed and used for free. (http://www.panelcheck.com). It can provide an intuitive and easy-to-use graphical user interface that handles all statistical computations in the background and visualises results in different types of plots (Tomic 2010). The initial version of the ConsumerCheck software, which was launched in 2016, was specifically designed to analyze four types of data that are common in sensory and consumer science, i.e. consumer liking data, consumer characteristics data, product design data and descriptive analysis/sensory profiling data. The ConsumerCheck project home page (http://www.consumercheck.co/) provides information about the software, links to downloads as well as installation instructions. This enables the practitioner and non-statistician to concentrate on performance analysis rather than spending time on trying to apply algorithms on data by themselves.

In face of the context stated, this research ending point is to analyze sensory profile and acceptance in 12 high-selling herbal beverages containing chrysanthemum samples from the Chinese online market. Using PanelCheck identified reliability of descriptive sensory analysis of the experimental results, panel members having good discrimination, and acquired information about the difficulty to recognize and evaluate some attributes. ConsumerCheck was used to analyze consumer preference data, and establish the correlation between descriptive sensory analysis and consumer preference. The research results can maximize the extraction of sensory attribute information of samples, thereby guiding improvements in sensory quality and product development. At the same time, the results also provided a simple and open-source software to data statistical method for practitioner without commercial software and programming skills.

2.1 Herbal beverages containing chrysanthemum samples

12 samples of herbal beverages containing chrysanthemum were purchased from https://www.taobao.com/, which is China's largest shopping website (Table 1). These samples have high-selling on the website. Original image of 12 samples were shown in Figure 1.

2.2. Sensory Analysis

40 mL samples of herbal beverages were presented in 50mL disposable plastic cups coded with a three-digit. All samples were evaluated at room temperature.

All the assessments were carried out in the Sensory Analysis Laboratory of the School of Biological and Food Engineering/Chuzhou University, designed in accordance with ISO 8589 (ISO 8589 2007), under artificial daylight and temperature control (22°C), in a separate area equipped with tables and chairs, specially conditioned for the test to enable participants to conduct the task comfortable and without distractions.

2.3 Descriptive sensory analysis

An expert panel evaluated the 12 samples by quantitative descriptive analysis (QDA). Judges with previous experience in descriptive analyses were preselected among food safety Department students. Finally 14 students were selected as assessors. The selection and training of the assessors followed ISO standard 8586:2012 for guidelines of selection, training, and monitoring of trained and expert sensory assessors (ISO 8586 2012). Basic taste identification and intensity ranking tests were conducted to measure the assessors’ sensitivity to different basic tastes of sweet, sour, salty, bitter and umami. The assessor was considered to pass the selection test if they scored a minimum of 80% for both identification and ranking intensity tests. The assessor with a score less than 80% will be further trained and asked to repeat the test until they are able to obtain 80%. Those who did not pass the second attempt will be excluded from the entire test. In our study, all assessors have passed the selection criteria. All assessors had received extensively training in profiling various types of foods containing chrysanthemum and specially on chrysanthemum tea. Assessors discussed all generated attributes and ordered a final list with 11 attributes, definitions and references (Table 2). Then assessors were further trained about attributes using the references within six 45 min sessions.

Appearance, aroma, flavor, texture of the samples were analyzed using on a 10-cm unstructured scale (from unnoticeable to very strong). 4 samples as a session were monadically presented using a balanced block design and assessed daily. Each sample repeated three times. Crackers and taste-free water were provided for palate cleansing.

2.4 Consumer acceptability test

A total of 75 consumers participated in the test (36 women and 39 men, aged 17－24 years). All were consumers of herbal beverages and declared no food allergies. All were recruited among the student population of the Chuzhou University. Participants signed an informed consent form and received two bottles of milk (worth 6RMB) as a small gift for their participation. The informed consent form explained the characteristics of the test and informed consumers that they were going to try herbal beverages containing chrysanthemum samples. Consumers were asked to evaluate their odour and taste liking using a 9-point hedonic scale, where 1= dislike extremely and 9= like extremely. The 12 samples were evaluated by consumers in a single session. Crackers and taste-free water were provided for palate cleansing.

2.5 Statistical analysis

The PanelCheck V 1.4.2 software was used to analyse sensory data. Panel performance was monitored with Line plot, Tucker-1 plot, Eggshell plot, Manhattan plots, and the test data were analysed using two-way ANOVA and Spider's web plot. The multivariate data analysis test named principal component analysis (PCA) was performed to identify relationships among samples and their sensory attributes.

The ConsumerCheck 2.3.1 software was used to analyse consumer liking date. Box plot and Stacked histogram visualise the distributions of the liking ratings across all consumers for each of the tested products. Preference mapping was obtained by partial least squares regression (PLSR).

3.1 Descriptive sensory analysis

3.1.1 Line Plots

Figure 2-1 shows the Line plots for each sample with all 11 attributes displayed with all scores and replicates of each assessor. The line connecting the attributes indicates the panel average scores for each attribute and therefore the plot visualizes the profile of each sample.

Figure 2-2 displays Line plots of A2, and A13 for sample 875. In each plot is shown the three repetitions scores of each attribute . In this way, it could be detect how each assessor has evaluated the sample. In Figure 2-2 it can be observe that A13 has a good reproducibility since replicates lie close to each other. A2 does not have a good reproducibility at “ Sweet Taste”, because the scores of replicates are far from each other. This type of results helps the panel leader to identify repeatability problems of single assessor ( Tomic et al. 2013).

3.2.2 2-way ANOVA

As a first step of the statistical elaboration, 2-way ANOVA was applied on the data sets of the panel. The results are visualized on Figure. 3. The plots have three different colors. The red means the significance level is p<0.001, the orange means p<0.01, the yellow means p<0.05 and the gray means the non-significance. The main reason for using this method is to eliminate the non-significant descriptors from the further analysis. The method is based on modeling samples, assessors and their interactions in two-way ANOVA model (Lanza et al. 2020). Here, our focus is primarily on the product effect. The non-significant attributes were excluded from the further analysis. According to this panel has one non-significant attributes , we have decided to exclude evaluation of “Honey Taste” by the subsequent data elaboration.

3.1.3 Tucker-1 plots

The multivariate analysis method Tucker-1 is applied in order to get an overview over assessor and panel performance using multiple attributes. The assessors in the inner ellipse for attributes meaning that they have more than 50% systematic variation of the variation explained by PC1 and PC2 (Attila et al. 2013). By screening through the plots (Figure.4-1), one can see that the overall performance between the 14 assessors can be considered as very good for “Sweet Taste”, “Astringent” and “Overly sweet taste”. Most of assessors have very high explained variance for “Yellowish-brown”, “Brillancy”, “Fragrant sweet flavour”, “Chinese herbal scent flavour”, “Honey taste”, “Lubrication” and “Astringent”. “Mint flavour” shows that 13 assessors are in the inner ellipse indicating that lower than 50% in PC1 and PC2. The consistency of the panel was poor on “Mint flavour”. For “Fragrant sweet flavour” it is obvious that panel A5and A6 disagrees with the other panels , since it is located alternate location of the other assessors. For “Lubrication” , A2 was disagreement of this assessor respect to the panel.

Figure.4-2 shows 14 Tucker-1 plots, one for each assessor. For example, the explained variances for attributes “Mint flavour”, “Lubrication” and “Chinese herbal scent flavour” of A2 are lower than 50% in PC1 and PC2. These results indicate a very strong disagreement of this assessor respect to the panel.

3.2.4 Manhattan plots

After screening through the Tucker-1 plots one may consult Manhattan plots (Figure. 5) for comparison of the systematic variation for a specific attribute across all assessors (Talsma et al. 2013). The Manhattan plots confirm what was shown in the Tucker-1 plots. The lighter a colour tone in a Manhattan plot is for a specific assessor-attribute combination, the more systematic variation is given (Naima et al. 2013). “Mint flavour” had the largest black area, need six principal components to reach a high level of explained variance. The color of “Sweet Taste”, “Astringent” and “Overly sweet taste”appeared very light gray indicated that majority assessors had a high percentage of explained variance already after two or three principal component. In“Lubrication”, it can easily seen that A2 differs from the other assessors. The lone dark bar indicated that A2 had less systematic variance for this attribute than the other assessors and need six principal components before explained variance was comparable with the other panels. The results further explained Tucker-1 plots (Figure. 4-1).

3.2.5 Eggshell plots

The Eggshell plot is a powerful graphical method that is valuable when it comes to detecting differences or disagreements between assessors. The panel leader gains a graphical over[1]view over all assessors in one plot allowing easy identification of those assessors that differ ranking-wise from the consensus (Tomic et al. 2007).

Figure 6 shows the Eggshell plots for the eleven attributes. The plot as the smooth line touching the consensus rank axis represents the consensus rank, while each of the other lines represent the cumulative ranking of one particular assessor. Profile plots show that the disagreement in evaluating the samples is strongest for the attributes “Mint flavour” as already observed in the Tucker-1 plots in Figure. 4-1 and Manhattan plots in Figure. 5. The Eggshell plot for “Yellowish-brown”, “Brillancy” and “Fragrant sweet flavour” also shows good agreement among all assessors, except for assessors A5 , A12 and A5, A6, respectively (that need to be retrained with standard references). The Eggshell plot for remaining descriptors show a more complex situation that could require the retraining of all assessors.

3.2.6 Sensory differences of investigated samples

The complete sensory profile of 12 samples is presented in Figure 7. Figure 8 shows the PCA performed using the means values of the attributes generated by QDA. Only descriptors that showed significant differences between samples were selected. A variability of 84.8% with two principal components was explained. The closer the samples are, the greater the similarity between samples. The closer the sensory attributes are, the greater the correlation between the attributes. The closer the sample is to the sensory attribute, the stronger that attribute becomes. The 12 samples were distributed in 4 different quadrants. Samples with 807, 244, 641, 765, 982, 636, 219, 975, 875 were positioned on the right side of PC1, while samples of 349, 468, 713 on the left side of the PC1. Samples with 641, 765, 982, 636, 219, 975 were characterized with brillancy, lubrication, fragrant sweet flavour, sweet taste, overly sweet taste, samples with 349 were mostly characterized with bitter taste and astringent, and sample with 875 were mostly characterized with Chinese herbal scent flavour.

3.2 Consumer acceptability test

Figure 9 shows the box plot for the the investigated samples consumer liking data where consumers rated 12 food products on a hedonic scale from 1 to 9, where 1 represents “dislike extremely”and 9 represent “like extremely”. The box plot shows that across all consumers sample with 875, 807, 713, 636, 982, 219, 244, 468 received the highest (9) and lowest (1) rate by at least one consumer, while sample with 975, 641, 349,765 received the highest (8) and lowest (1) rate by at least one consumer. The green boxes for each line visualise the distribution of the ratings between the 25th and 75th percentile. The dark green line across the green boxes shows the median rating for that product. Overall, samples 875 and 219 were liked most, whereas samples 713, 636, 349 were liked most least.

In Figure 10 a stacked histogram plot is shown for the same data as presented earlier in a box plot in Figure 9. Stacked histograms provide another and richer way of visualising consumer liking data. With the stacked histogram each bar represents one sample and each colour in the bar represents a certain rating of the product. For sample 807 one can see that 8 consumers or 10% of the total number of consumers rated this sample with 1 (dislike extremely). 14 consumers or 18% of the consumers rated sample 807 with 5, and so on. In this way the distribution of the ratings is visualised in a more detailed way than in the box plots.

3.3 Preference mapping

Preference mapping is applied simultaneously to consumer liking data and descriptive analysis data. The data set consists of both QDA data and consumer liking of the same samples. Only the averages will be considered for QDA (Ns et al. 2012). The aim is to find drivers of liking that may determine why some products are preferred over other (Stavros et al. 2015). In Consumer Check , two standard statistical tools applied to build a preference mapping model are partial least squares regression (PLSR) and principal component regression (PCR). Internal preference mapping used consumer liking data as the X matrix and QDA data as the Y matrix. Since the distribution of the products is influenced by the common variance in both X and Y matrix, PLSR was chosen to compute the preference model.

Figure 11 shows the X scores of the preference mapping model visualising how the products relate to each other in the space. As can be seen from Figure11 PC1 and PC2 explain16% and32% (first number in parenthesis) of the variance in the X matrix(consumer liking). PC1 and PC2 explain 53% and 12% (second number in parenthesis) of the data in Y (QDA) totalling 65%. Similar products are located close to each other and dissimilar products have a larger distance between them spanned by the first two components. Sample 713, 349 , 765 and 636 are on the opposite side of other samples with regard to PC1, they are some difference with PCA where analysis was based on QDA data only (Figure 8).

Figure 12 shows the actual preference map that is used for interpretation and visualisation of consumer preferences and drivers of liking. In this plot both the correlation loadings from X and Y are displayed in the same plot. Consumers in the inner circle closer to the origo don't discriminate between the products with regard to the variation described by PC1 and PC2. The closer the attribute is to the outer ring, the higher the variances that can be explained. Many consumers prefer products with high intensities of fragrant sweet flavour, brillancy, lubrication, sweet taste, overly sweet taste (lower left part of the plot) since a large part of the consumers are in proximity of those attributes. And these attributes are highly correlated. Chinese herbal scent flavour, mint flavour, yellowish-brown, bitter taste and astringent are less preferred although there are a few consumers that prefer high intensities of these sensory attributes. Chinese herbal scent flavour and yellowish-brown, bitter taste , astringent are also correlated, however to a lesser degree.

In face of the context stated, this research ending point is to analyze sensory profile and acceptance in 12 high-selling herbal beverages containing chrysanthemum samples from the Chinese online market. Based on QDA, a total of 11 sensory attributes were generated. The main characteristic of herbal beverages containing chrysanthemum can relate to visual appearance of yellowish-brown and brillancy , while the aroma is characterised by fragrant sweet, Chinese herbal scent and mint . The flavour is represented by sweet, bitter and honey in addition to lubrication , astringent and overly sweet taste. The panel members were evaluated one by one applying the univariate and multivariate (two-way ANOVA, Line plot, Tucker-1 plot, Eggshell plot, Manhattan plots, Spider's web plot, PCA) statistical techniques by the open-source software (PanelCheck V 1.4.2 Software, 2015). The software can quickly identify underperforming assessors by displaying performance indices in a table or plotting them in a graph. Mint flavour required additional training since the panelists were not in full consensus for some attributes. However, the reliability of the test is considered to be good based on test repeatability. ConsumerCheck (ConsumerCheck 2.3.1 Software, 2020) is also an open-source data analysis software tailored for analysis of sensory and consumer data. Besides some simple descriptive statistics the main statistical methods implemented in ConsumerCheck are principal component analysis, preference mapping, partial least squares regression, principal component regression and conjoint analysis. The software can be able to provide preference mapping that a correlation with consumer hedonic scores represented in a perceptual map, which should be understood as the acceptance drivers of herbal beverages containing chrysanthemum samples. The study gives insight to producers into which sensory characters are positive and which are negative influences on liking. The results showed that the attributes that mainly drive consumers' preferences were fragrant sweet flavour, brillancy, lubrication, sweet taste, overly sweet taste.

Thus, this study could guide the development of Chinese traditional chrysanthemum drink. At the same time, the results also provided a simple and open-source software to data statistical method for practitioner without commercial software and programming skills.

Compliance with ethical standards

Since the research involved tasting and food quality evaluation, the academic committee of the School of Biological and Food Engineering of Chuzhou University approved the study and exempted it from review. According to the FDA, CFR－Code of Federal Regulations Title 21, food quality and consumer acceptance studies are exempt from review ( US Food and Drug Administration 2017).

Author Contribution

Long Men wrote the main manuscript text. All authors reviewed the manuscript and participated in the experiment.

Data availability statement

All relevant data are within the paper.

Conflict of interest statement

We declare that we have no conflict of interest.

Acknowledgements

This work was supported by Anhui Education Department (2022AH051120) and CHUZHOU University (2022XYJB31, 2020jyz027, 2022ghjc07, 2022hqkc001, 2022zsxm007, 2023zckc007) .

A González-Mohino, A Jiménez, Rufo M , et al. (2021) Correlation analysis between acoustic and sensory technique data for cooked pork loin samples. LWT- Food Science and Technology, 141:110882.
Ayustaningwarno F , Fogliano V , Verkerk R , et al. (2021) Surface color distribution analysis by computer vision compared to sensory testing: Vacuum fried fruits as a case study. Food Research International, 143:110230.
Attila G , S Dániel, F Tamás, et al.(2013) Statistical possibilities of the of the Panelcheck software at monitoring of sensory panel performance. Elelmiszervizsgalati Kozlemenyek, 59(1):15-27.
Bae S M , Lee S C . (2008) Effect of Hot-air Drying Temperature on Volatile Compounds in Chrysanthemum boreale M. Flowers. Korean Journal of Food Science & Technology, 40(4):466-469.
Bhardwaj A , Sharma N , Sharma V , et al. (2022(Pt.B)) Assessing the consumer acceptance and storability of chitosan and beeswax coated cellulose packaging for whole wheat bread. Food Control, 133:108682.
Campoceballos D A , CA Gavirial-López. (2023) Optimising consumer acceptability of Cauca specialty coffee through roasting profiles on acidity and body sensory attributes. Czech Journal of Food Sciences, 41(02): 64-72.
Deng Z , Fu Z , Yan W , et al. (2021) The Different Effects of Chinese Herb Solid Drink and Lactulose on Gut Microbiota in Rats with Slow Transit Constipation Induced by Compound Diphenoxylate. Food Research International, 143:110273.
Faccioli L S , Klein M P , Borges G R , et al. (2021) Development of crackers with the addition of olive leaf flour (Olea europaea L.): Chemical and sensory characterization. LWT- Food Science and Technology, 141(264):110848.
Huang X , Hsieh F . (2010) Physical Properties, Sensory Attributes, and Consumer Preference of Pear Fruit Leather. Journal of Food science, 70(3):177-186.
Ho P . (2015) Statistical methods and tools for analysing sensory food texture. Modifying Food Texture, 20:45-87.
Jo Y , Kim M K . (2020) Influences of appearance characteristics on consumer acceptance and perception of "gu-soo" in fermented soybean paste (doenjang). Journal of Sensory Studies,2020,35(05):e12597.
ISO, ISO 8589. (2007) Sensory analysis - General guidance for the design of test rooms.
ISO, ISO 8586. (2012) Guideline for selection, training and monitoring of selected assessors and expert sensory assessors in General guidelines for the selection, training and monitoring of selected assessors and expert sensory assessors.
Jérme Pagès, Husson F .(2013) Multiple factor analysis: Presentation of the method using sensory data. John Wiley & Sons, Ltd.
Luo D , Chen J , Gao L , et al. (2017) Geographical origin identification and quality control of Chinese chrysanthemum flower teas using gas chromatography–mass spectrometry and olfactometry and electronic nose combined with principal component analysis. International Journal of Food Science & Technology, 52(3):714-723.
Liao Xiaoqin, Zhang Xingsu, Li Xin, et al. (2023) Lysine malonylation of DgnsLIPID TRANSFER PROTEIN1 at the K81 site improves cold resistance in chrysanthemum. Plant physiology, 163: 114818.
Li Xiang, Li Rongxia,Wang Xueyu, et al. (2023) Effects and mechanism of action of Chrysanthemum morifolium (Jinsi Huangju) on hyperlipidemia and non-alcoholic fatty liver disease. European Journal of Medicinal Chemistry, 255: 115391.
Lattey K A , Bramley B R , Francis I L . (2010) Consumer acceptability, sensory properties and expert quality judgements of Australian Cabernet Sauvignon and Shiraz wines. Australian Journal of Grape & Wine Research, 16(1):189-202.
Lanza B , Amoruso F . (2020) Panel performance, discrimination power of descriptors, and sensory characterization of table olive samples. Journal of Sensory Studies, 35(1): e12542.
Mauricio A A , Palazzo A B , Caselato V M , et al. (2016) Generalized Procrustes Analysis and External Preference Map Used to Consumer Drivers of Diet Gluten Free Product. Food and Nutrition Sciences, 7: 711-723.
Mudasser N M , Zameer D A , Omer N , et al. (2021) Variations in growth performance, meat quality and consumer preferences influenced by difference of breed in buffalo bulls [J]. Animal Production Science Anim. Prod. Sci., 61(16):1715-1722.
Martins Z E , Pinho O , Ferreira I M P L O . (2017) Fortification of Wheat Bread with Agroindustry By-Products: Statistical Methods for Sensory Preference Evaluation and Correlation with Color and Crumb Structure. Journal of Food Science, 82(7-9):2183-2191.
Naima H A , Mehdi S , Morched C . (2013) Training and Control of Performance of a Tactile Sensory Panel. Journal of Applied Sciences, 13(3):366-376.
Ns T , Tomic O , Endrizzi I ,et al. (2021) Principal components analysis of descriptive sensory data: Reflections, challenges, and suggestions.Journal of Sensory Studies, 36(5):12692.
Ubeda C , Callejon R M , Troncoso A M , et al. (2017) Consumer acceptance of new strawberry vinegars by preference mapping. International Journal of Food Properties, 20(9-12):2760-2771.
US Food and Drug Administration .(2017). CFR-Code of Federal.
Stavros P , Oliver R A , Helen D ,et al. (2015) Preferred reporting items for studies mapping onto preference-based outcome measures: the MAPS statement[J].International journal of technology assessment in health care, 31(4):230-235..
Tzompa-Sosa D A , Dewettinck K , Gellynck X , et al. (2022) Consumer acceptance towards potato chips fried in yellow mealworm oil. Food Quality and Preference, 97:104487.
Tomic O , Luciano G , Nilsen A , et al. (2010) Analysing sensory panel performance in a proficiency test using the PanelCheck software. European Food Research and Technology, 230(3):497-511.
Tomic O , Forde C , Delahunty C , et al. (2013) Performance indices in descriptive sensory analysis – A complimentary screening tool for assessor and panel performance. Food Quality & Preference, 28(1):122-133.
Talsma P . (2016) Assessing sensory panel performance using generalizability theory. Food Quality and Preference, 47:3-9.
Tomic O , Nilsen A , Martens M , et al. (2007) Visualization of sensory profiling data for performance monitoring. LWT - Food Science and Technology, 40(2):262-269.
Yuan J , Hao L J , Gang W , et al. (2015) Effects of drying methods on the phytochemicals contents and antioxidant properties of chrysanthemum flower heads harvested at two developmental stages. Journal of Functional Foods, 19:786-795.
Zhan Ge, Long Men, Shan Kai, et al. (2022) Antioxidant Effect of Chrysanthemum morifolium (Chuju) Extract on H₂O₂-Treated L-O₂ Cells as Revealed by LC/MS-Based Metabolic Profiling. Antioxidants,11(6):1068.

Table 1 Information of 12 samples

No.	Product Name	Manufacturer	code
1	Original chrysanthemum plant drink	China Resources C'estbon Beverage (China) Co. Ltd.	875
2	Chrysanthemum plant drink	Shenzhen Vita Food and Beverage Co., LTD	975
3	Honey chrysanthemum tea	Xiamen Huierkang Food Co., LTD	807
4	Chrysanthemum plant drink	Shenzhen Shenhui Food Co., LTD	641
5	Puer chrysanthemum tea	Dongyang Beverage (Changshu) Co., LTD	713
6	Chrysanthemum and wolfberry tea	Li Shi Industrial Development Co., LTD	636
7	Fiber tea-Hang white chrysanthemum tea	Healing Forest (Tianjin) Beverage Co., LTD	349
8	Chrysanthemum tea	Anhui Dongpeng Food and Beverage Co., LTD	765
9	Chrysanthemum plant drink	Xiamen Huierkang Food Co., LTD	982
10	Sunshine chrysanthemum tea	Coca-Cola Bottler Production (Dongguan) Co., LTD	219
10	Chrysanthemum plant drink	Henan Zhongwo Industrial Co. LTD	244
12	Chrysanthemum and wolfberry tea	Zhangzhou Dingjin Food Co., LTD	468

Table 2 Definitions of the attributes used by the expert panel to describe the samples.

Descriptors	Definitions	References
Yellowish-brown	Appearance color of the drink	Royal Horticultural Society (RHS) color chart containing 920 different colors^[26]
Brillancy	Appearance glossiness of the drink	Master Kong Iced Black Tea
Fragrant sweet flavour	Aromatic note resembling ripe fruits	Pulp mixture of strawberries, blueberries and raspberries
Chinese herbal scent flavour	Aromatic note resembling dried Chinese herbs	2g of Chinese traditional twenty-four flavor herbal tea bags ( Mixture of dried honeysuckle, Houttuynia Cordata, gardenia, dandelion, light bamboo leaves, mint, purslane, chrysanthemum, lotus leaves, pueraria, mulberry leaves, Elsholtzia patchouli and fresh reed root, white grass root）
Mint flavour	Aromatic note resembling Mint，pungent smell with a cool feeling	2g of freshly crushed mint leaves
Sweet Taste	Taste of sugar	6% sucrose solution
Bitter Taste	Taste of substances such as caffeine or quinine	0.08% caffeine solution
Honey Taste	Flavor sensation resembling honey	3% honey water solution
Lubrication	Taste of finesse, moist and smooth	Healing Forest's original milk tea of low -sugar and fat
Astringent	Taste of tea, tannin, dry-mouthfeel sensation	0.1% alum solution
Overly sweet taste	Unpleasant sensation results from an excessive amount of sweetness	12% sucrose solution

No competing interests reported.

Download PDF

Editorial decision: Major revision
21 Oct, 2023
Reviews received at journal
29 Aug, 2023
Reviewers agreed at journal
26 Aug, 2023
Reviewers invited by journal
26 Aug, 2023
Submission checks completed at journal
21 Jun, 2023
Editor assigned by journal
21 Jun, 2023
First submitted to journal
20 Jun, 2023

You are reading this latest preprint version

Panel performance, sensory characterization and consumer preference of Chinese traditional chrysanthemum drink

Status:

Version 1

Abstract

Figures

1 Introduction

2 Materials and Methods

3 Results and discussion

4. Conclusions

Declarations

References

Tables

Additional Declarations

Status:

Version 1