| Today, there is little doubt in the wine industry that personnel, who have been trained to manage sensory evaluation of wine at their respective cellars, add great value to the winery’s quality control activities, as well as aid with product development. The purpose of this article is to provide a short and graphic overview of the main wine sensory evaluation methods, as well as some guidelines for people in the industry, on what factors should be taken into account when choosing an appropriate method. This article (Part 1), is the first in a series of four and provides information about Descriptive analysis (DA), the classic method where each wine is profiled in detail, as well as information about alternative rapid methods. Part 2 focusses on the Sorting method, Part 3 on Projective mapping and Napping, and Part 4 on the Check-all-that-apply (CATA) method. This series is concluded in Part 4 by a summary where the methods are compared in terms of their practical requirements. These include the number of panellists required, time required for the completion of the tasting by the panel and difficulty of the tasks, amongst others. On the research side of sensory evaluation, the main focus currently being investigated worldwide (in food and wine), is methodology. The question: “What is the best method to do sensory tests on wine?” is often asked by the industry. Time is money, and it is no wonder that the current objectives of sensory research are to find cost-effective and less time consuming methods that are sustainable in the long term. Development of new sensory methods, as well as optimisation of existing ones, is also one of the focus areas of the sensory research currently being done at the Institute for Wine Biotechnology and Department of Viticulture and Oenology (IWBT-DVO) and the Institute for Grape and Wine Sciences (IGWS), Stellenbosch University (SU). When we speak of sensory testing methods this includes not only the evaluation of aroma and taste of the wine, but also the handling and statistical analysis of the measurements, or data. One of the main aims is to make the statistical procedures as user friendly as possible and to present the information in simple graphic ways, so that it can easily be interpreted. In short, the expression fit-for-purpose is especially applicable to wine sensory evaluation. There is a myriad of methods in literature that are available for the ordinary user, a situation that can become very confusing to interpret. |  |
| [wds id=”64″] |
Compiling a panel
With sensory analysis the “instrument” is a human panellist. Different panels can be formed by people with different levels of knowledge and experience of the product to be tested. The following three types of panels are generally used:
- Ordinary (non-product expert) consumers;
- Wine experts (e.g. winemakers or experts from the wine industry); and
- Trained professional sensory tasters (not necessarily wine experts).
Methods for sensory evaluation
In terms of the method, the whole set of wines can be profiled in depth, capturing intensity ratings for each characteristic on a line scale. Other more rapid methods are based on only evaluating similarities and differences between wines in a set. It is important to choose the right method, and factors to be considered include the number of panellists, their availability, and type of panel that is required, the method to be used and choice of data analysis techniques. With data analysis it is important to consider which analysis you can do yourself and, which methods of analyses will require support from experts. The most important aspect is that the exact purpose of the sensory evaluation be clarified, this makes the choice of panel and test method easier. Some of the useful methods that can be evaluated for their efficiency by the industry are discussed in the following sections, as well as in the series of three articles that follow on this review.
Descriptive analysis for complete sensory profiling of wine
Descriptive analysis (DA) is the most well-known and regularly used method of sensory analysis. DA is used to identify sensory characteristics of a product and score their intensities. A well-trained panel is essential for quality results. The standard procedure requires that a panel of between eight and 12 judges is trained to identify and score the intensities of aromas and tastes. For this reason no more than eight to 10 wines are normally included per test. Panel training is followed by the testing stage, in which the same set of wines used for training is scored by the panel. Data generated with DA is numerical (intensities can for instance be rated as a number out of a total of 100), as well as descriptive, and can be easily combined with chemical measurements on the same set of wines. This would be desirable, for example, when correlations between chemical and sensory data are investigated. DA is the method of choice when detailed sensory information of the products is required, when a comprehensive list of all the attributes of a single product is required, or when one wants to quantify specific taste/aroma differences between a few products.
Although DA is the most well-known and widely used method for sensory analysis, it is a lengthy and costly process. Training of a panel can take between 10 and 30 hours depending on how complex the wine is, and how many samples you wish to profile.
Rapid sensory analysis methods
It is also more than likely that one would wish to profile a considerable number of wines, as quickly and as accurately as possible. To fulfil this need for faster/less expensive sensory analysis, came the development of alternative more rapid sensory analysis methods. These methods still use panellists, but can be completed in a short period (in some instances, one hour is sufficient). Several of the rapid methods used in the food industry today were not initially developed for wine and the sensory facility of the IWBT-DWW and IGWS, SU, started a Winetech-funded research project a few years ago to test the efficiency of the rapid methods when applied to wine. Where necessary, methods were adjusted to improve their performance so that the industry can use them. Throughout this series of papers, we use the results of our research to make recommendations regarding applications to wine.
As can be seen in Figure 1, the rapid methods can be divided into two broad groups, namely:
- Verbal-based methods (descriptions); and
- Similarity-based methods (grouping).
Within these two groups they can be further split into:
- Flash profiling (FP);
- Check-all-that-apply (CATA);
- Sorting; and
- Projective mapping (PM).
Within the last four mentioned, there are a number of variations that differ in their degree of difficulty and complexity.
Free choice and Flash profiling (FP)
With Free choice profiling (Figure 2), panellists use their own words to describe a wine sample, as the name suggests. There is no training and no list of descriptors is given. The follow on from that is called Flash profiling (Figure 2), where panellists, using the list of descriptors they generated, then rank each wine sample in terms of how strongly they perceive each descriptor in each wine. FP is a good method to use when one needs to rapidly identify a product set’s most important attributes. Free choice and Flash profiling are good methods to use to quickly identify the most important descriptors in a set of wines.
Check-all-that-apply (CATA)
With the Check-all-that-apply method the panellist is given the samples, as well as a list of descriptors or phrases (Figure 3). The panellist is then required to smell/taste the sample and choose descriptors that they feel most accurately describe the wine. For this method, if consumers are used, words or phrases such as ‘like it’, ‘I would buy it’ and ‘I don’t like it’ can be included. From this CATA method, a variation was recently added to the portfolio of wine sensory methods and referred to as the attribute frequency of citation method. This modification requires that panellists are trained on the usage of the descriptors on the list, with the aim of improving consensus regarding the meaning of words and simplifying data analysis.
When setting up a CATA list, the aroma lists available for each cultivar are good starting points, and other attributes can be added at the panel leaders’ discretion. One of the most important factors to keep in mind when setting up a CATA list is that the order which the terms appear in, can affect the results. Terms at the top of the list are used more often than those at the bottom. To combat this problem, the following recommendations can be kept in mind:
- The order of the terms on the list should be changed for each assessor;
- Grouping of terms and thereby shortening of the list; and
- Making use of multiple shorter lists.
CATA lists can also be used in combination with other sensory tests like Sorting or Mapping. Figure 3 is an example of the full general wine aroma CATA list originally developed by Dominique Valentin (AgroSup, Dijon, France) and adapted by the sensory laboratory at IWBT-DVO and IGWS, SU, to suit the South African wines and requirements of industry testing.
Sorting
Sorting is a similarity-based test (Figure 4), where samples are grouped according to similarities or differences. Each panellist sorts their set of samples into groups, on a sheet of paper for instance, in a way that makes sense to him/her, and then provides a list of attributes to describe each of their groups. All samples are presented simultaneously, and this technique can be used to evaluate aroma, taste or visual appearance. There are two simple rules, which must be followed; firstly the panellist must make more than one group and secondly, he/she cannot put each sample in its own group. Sorting is fast (normally it can be completed within one hour) and can be done by trained or inexperienced panellists, because it does not require consensus from panellists and there is no quantitative rating. Sorting can also be done with a large number of samples, ideally between nine and 20.
Projective mapping (PM) and Napping
For Projective mapping (Figure 5), also referred to as Napping, samples are sorted/mapped on a large sheet of paper (usually size A2) in two dimensions. The name Napping comes from when it was originally done on a tablecloth (nappe in French) instead of paper. Panellists are instructed to arrange each sample on the page according to their perceived sensory similarities and differences. The more similar the samples, the closer together they would be placed, and the more different, the further apart the samples would be placed. The positions of the glasses are then marked on the paper and using the bottom left hand corner of the page as the origin, X and Y co-ordinates are obtained for each sample. The distances (in cm) between the glasses represent their similarities and differences. Panellists also write three to five descriptors for each sample or cluster of samples. This method is fast and can be completed within one hour. It is easy for either ordinary consumers or trained panellists to complete.
Polarised projective mapping (PPM)
Polarised projective mapping (Figure 6) is a similar process to Mapping, there is just one difference. One wine sample is defined as a pole. This pole is used as a reference and the other samples are arranged (mapped) in relation to it. A preliminary PM can be done to identify potential poles/reference samples, or alternatively the poles can be chosen by the panel leader.
Which method is the best?
DA is the method of choice when:
- Detailed sensory information of a product is required;
- When a list of all the characteristics observed by the panel, including their intensities, for each wine sample is required; and
- When specific taste or aroma differences between a couple of products needs to be quantified. Software programmes like Compusense are available and are convenient for capturing DA testing data. Although DA is the most commonly used method for sensory evaluation, it is a long and costly process.
The rapid methods do however also have limitations, and the choice of method depends on the size of the panel, the timeframe and budget that is available. For the results to be statistically valid for the rapid methods, you need 30 individual answers (or 15 answers in duplicate). Data obtained from DA is numeric, as well as descriptive (whereas for the more rapid methods the data is mostly descriptive), and can therefore be combined with other numerical measurements from the same set of wine, e.g. consumer liking or chemical data. Mapping however, because of the co-ordinate data it produces, has the potential to be combined with other data. This however is still being worked on.
With Flash profiling, CATA and DA all the wine samples are profiled individually, therefore if wines are added or removed, this will make no difference to your results. In contrast, Sorting and PM is a holistic approach where all the samples are compared to one another in order to be grouped or arranged on the tasting sheet. Due to the fact that the wine samples are grouped in relation to each other, adding or removing a sample from the set could alter the results. With CATA and Sorting methods, you will generally only pick up large differences between wines, and be able to identify the most dominant characteristics. These methods are limited in their ability to differentiate between samples where differences are intensity-based.
Overall, the advantages of the rapid methods are that they require less panel training as the panel does not have to come to consensus on scaling of descriptors, as is the case with DA. The rapid methods are definitely faster for the panel to complete, but not for the panel leader/analyst and more research is currently being done to speed up the data capturing and subsequent data analysis stages.
The following articles in this series will discuss three rapid methods for sensory evaluation of wine, namely Sorting (Part 2), Projective mapping (Part 3) and CATA (Part 4). These articles will supply more detailed explanations and illustrations, as well as guidelines on how to choose the best method for your task in the cellar environment.
Acknowledgements
This research was funded by Winetech, project IWBT W13/02: “Rapid descriptive sensory methods for wine evaluation – special focus on further optimisation of rapid methods and streamlining of workflow”. The Institute for Grape and Wine Sciences (IGWS), Stellenbosch University, NRF and THRIP are acknowledged for financial support. Dominique Valentin, AgroSup, Dijon, France is acknowledged for the research on the CATA method.
References
Cartier, R., Rytz, A. & Lecomte, A. et al., 2006. Sorting procedure as an alternative to quantitative descriptive analysis to obtain a product sensory map. Food Quality and Preference 17, 562 – 571.
Lawless, H.T. & Heymann, H., 2010. Sensory evaluation of food: Principles and practices. Springer, New York.
Meligaard, M., Civille, C.V. & Carr, B.T., 2006. Sensory evaluation techniques, 4th. CRC, Boca Raton, FL.
Risvik, E., McEwan, J.A. & Redbotten, M., 1997. Evaluation of Sensory profiling and Projective mapping data. Food Quality and Preference 8, 63 – 71.
Valentin, D., Chollet, S., Lelievre, M. & Abdi, H., 2012. Quick and dirty but still pretty good: A review of new descriptive methods in food science. International Journal of Food Science and Technology 47, 1563 – 1578.
Varela, P. & Ares, G., 2012. Sensory profiling, the blurred line between sensory and consumer science. A review of novel methods for product characterization. Food Research International 48, 893 – 908.
Consultation services offered by the Wine Sensory Facility, IWBT-DVO, Stellenbosch University:
- Selection of appropriate sensory methods.
- Guidance with selection, training, testing of panellists.
- Experimental design and practical workflow.
- Statistical analysis and interpretation of data.
– For more information, contact Carla Weightman at cweightman@sun.ac.za, Jeanne Brand at jeanne@sun.ac.za, Hélѐne Nieuwoudt at hhn@sun.ac.za or contact the authors at winwynsen@sun.ac.za.
