By Muhammad Salman Chukkan* & Ria Bhadra**
Sensory Evaluation is a commonly used term in the food industry and even for those who aspire to enter this industry. Yet, sometimes it is misunderstood by the professionals who are in this field of study. This misunderstanding essentially comes due to the lack of knowledge about this working body. According to Sensory Food Network Ireland, sensory evaluation means a scientific discipline that is used to measure and analyse people’s responses to products as perceived through the 5 senses namely sight, smell, touch, taste, and sound.
Sensory Evaluation is often conducted in the Food and Beverage industry due to its ability to foresee the pitfalls in the product itself. This helps companies introduce suitable and the best product to the market. It also helps to prevent companies from introducing a product that might not match consumer’s preferences with the help of sensory evaluation. Companies at times fail to understand the best possible manner of using this regulation.
The Bureau of Indian Standards have developed guidelines of sensory evaluation through the standard IS 6273. These standards come in three different parts-
- Part 1-Optimum requirement,
- Part 2- Methods and Evaluation Cards
- Part 3-Statistical analysis of data.
This article offers a basic overview of the standard.
Part-1 Optimum Requirements:
Part 1 of the standard IS 6273, deals with optimum requirements for the evaluation. This includes Personnel, selection of sensory panel, set up of laboratory, sample preparation and presentation etc. All the important aspects are discussed below:
Personnel:
1. Panel Organizer: A Panel organiser is the person who is in-charge of conducting a session. The personnel should form apt questionnaires by consulting the product developer and should be responsible for moderating, organising sessions, and act as a link between the product developer and the panel. No panel organizer shall impose their ideas on the panel lists.
2. Statistician: A statistician should choose the best suitable for sampling procedure, evaluation card and analysis.
3. Preparation Room i/c: The preparation room IC should make sure that all the sample preparation and presentation is decided by the statistician.
Panel:
A panel member’s selection is based on some basic criteria like their health, average sensitivity, capability of independent judgement, freedom from prejudice, and the member’s willingness to spend time during the course of action.
Depending upon the types of panel members chosen for a particular study by a sensory session organizer, the number of members may vary (as mentioned below):
1. Trained Panel: A small group of 5-10 members can be chosen and maintained to understand the intensity or the differences to determine the overall quality of food. They need not be experts in the panel. They can undergo training sessions and gradually develop the skill to analyse the products more efficiently. Especially in flavour industries, there is a need to undergo a higher degree of training in order to effectively analyse the broad spectrum of flavours.
2. Semi-Trained Panel: This panel should consist of members who are normally familiar with different classes of foods. They shall be able to discriminate and effectively communicate the differences. This panel must be used to understand the acceptability or preference of a final product. It is used before conducting full-fledged large-scale consumer trials. A typical group of semi-trained panels may consist of 25-30 members.
3. Untrained Panel: Untrained panel is also known as a consumer panel. This panel consists of randomly selected potential customers of the target markets. The panel members should be such that they ensure representation of the market area, from different age groups, sex, race and economic strata of a market. The group consists of minimum 100 members.
Laboratory setup:
There is a proper degree to how the sensory laboratory has to be designed. The sensory evaluation has to be conducted in a quiet, properly illuminated and odour-free room. The sensory booth has to be comfortable for the s to be able to carry out a prolonged analysis, whenever required. Temperature of the room, availability of drinking water, space and vicinity of wash basins are few of such requirements. The sample preparation room should be separate and the layout should be made such that a single person can efficiently attend to all the s in their respective booths.
Sampling and Testing:
The sampling should be carried out by a trained and experienced person, to obtain a representative sample. The samples are presented uniformly so that there is no involvement of biases. In case of foods which trigger intense physiological perceptions like vinegar, spices, high heat containing foods. There may be a need for dilutions during the sample preparation.
In addition to these, there are certain other etiquettes and rules of a sensory test as noted down below:
1. Temperature of sample: The temperature should preferably be at room temperature for difference testing. For preference testing it is suggested to serve at the common serving temperature of food.
2. Time of testing: The time for testing should be one hour before or after lunch timings.
3. Prohibition: There is a clear prohibition on the usage of pan, tobacco or alcohol at least half an hour before the test.
4. Number of samples: The number depends on how critical is the parameter and the profile of food is sustaining. If the sample is mild, then a large number of samples can be tested. In case the parameter profile is strong, only a few samples should be analysed at a time.
5. Coding: Coding can be done by using a table of random numbers (A random number table is a series of digits (0 to 9) arranged randomly in rows and columns).
6. Order of presentation: When a test involves multiple samples, the order of presentation of these samples is very essential. For example, a milder product should be evaluated first followed by a spicy product.
7. Evaluation Card: The evaluation card should be printed clearly and should be easily comprehensible to the panel members. It should not create any confusion.
Part 2 & Part 3 can be referred in Table A presented below:
Table A:
| S.No | Name of the Test | Statistical Test | Application |
| Difference Test: Difference testing is used to determine if foods differ in certain aspects. Some of these aspects include the test but are not limited to odour, taste, and texture. | The qualitative & quantitative differences and preferences are tested as per the evaluations resulting from the test product and that done by the . | ||
| 1 | Paired-Comparison test: Pair of samples are arranged and panelists are asked to find whether the samples are different (Simple difference test) or does it indicate which of the sample is less in specific sensory attribute. (Directional difference test). | X2 test or t-test is conducted. | Simple and directional differences for a specific parameter is measured. It can also be used for panel training. |
| 2 | Duo-trio test: Three samples are given out – out of which two is identical. The identical can be sample or reference.The panelists are asked to take the sample which is similar to the reference. | X2 test or t-test. | Different product samples are evaluated under the same quality parameter with a reference. It is used only when inter sample flavour has a carry-over effects that defines absence or quantity. |
| 3 | Triangle test: Two identical (reference) and one different (test) samples are given then panelists are asked to pick the odd one out. | X2 test or t-test. | To detect inter sample difference or panel testing difference. |
| 4 | Ranking test: All the samples are coded and ranked based on the intensity of an attribute or overall acceptability. | For a panel agreement coefficient of concordance is used if number of samples exceeds 7 use X2 test. If number is within 20, use rank sum analysis for preference/difference or else we can use X2. |
This is used to rank several samples of an attribute (intensity ranking) or preference of products (preference ranking) |
| Rating test: Difference tests is carried out with a qualitative aspect. | |||
| 5 | Monadic Test: Ask the panelist to identify the presence, absence and/or intensity of a specific quality parameter of a sample. Repeat the test at different times and compare the results | Convert intensity scores in numericals and ANOVA. | When a sample has a strong flavour that precludes the ability to test different samples in same session. |
| 6 | Two Stage Triangle test: Three coded samples are used – two identical (reference) and one test sample . The panelist is asked to pick the odd one out. Once it is done, then the panelist is asked to indicate the intensity of difference using scores to indicate the preference between the identical and odd sample. | Odd samples should be analyzed by binomial distribution. Discard the data if it points out wrong odd samples. The difference scores is analyzed by ANOVA. | Similar application of triangle test is carried and results in differences and preferences. |
| 7 | Hedonic Rating: These are asked to rate the intensity or acceptability on 7 or 9 pointer scale which is described by words. | Fast method: Convert rating to ranks and rank-sum analysis. Rigour method: Convert ratings to scores and ANOVA. |
Degree of pleasure or displeasure of a product. |
| 8 | Numerical Scoring test: Multiple sample are analyzed together. The scoring is usually done on 10 pointer scale or represented by a specific band of number for acceptability ( Example:1-3 Poor, 4-5 Fair, 6-8 Good and 9-10 Excellent) | t-test is done to compare two samples. Dunnet’s test for comparing against reference. Range method is used for multiple sample comparison and Duncan’s multiple range test is done for multiple comparison. | This is used to assess quality of a set of similar products, NPD. |
| 9 | Composite Scoring test: Weighted rating scale for each individual attributes based on their relative importance contributing to the overall quality of the product. Ask the panelist to rate each attributes based on the assigned weighted score and by adding all the individual scores. | t-test is used to compare two samples. Dunnet’s test compares against reference. Range method compares for multiple comparisons and then Duncan’s multiple range test is used for multiple comparison. | Product comparison and overall quality of grading is carried by s. |
| Please refer the standard, IS 6273 Part II for specimen Evaluation card for different evaluation method. | |||
Hence, we conclude that the art of sensory evaluation does not limit to mere testing, rather it involves hard-core science, psychology or statistics behind it. A lot of advanced testing methods and practices involve the use of software to carry out the statistical analysis part. With increasing popularity of sensory science in the food industry, there would be a need of more awareness in this area. By strictly following the standards and the guidelines, the efforts conducted during a sensory evaluation could be made more useful.
References
1) Guide for sensory evaluation of foods: Part 3 statistical analysis of data: Sec 1 difference/preference tests (First Revision)- (IS 6273 : Part 3 : Sec 1 : 1983)
2) IS 6273-2: Guide for Sensory Evaluation of Foods, Part II: Methods and Evaluation Cards
3)IS 6273-3-1: Guide for sensory evaluation of foods: Part 3 statistical analysis of data: Sec 1 difference/preference tests (First Revision)
4) IS 6273-3-2 Guide for Sensory Evaluation of Foods, Part 3: Statistical Analysis of Data, Section 2: Ranking and Scoring Tests
*MTech (Food Engineering and Technology)
Institute of Chemical Technology
**Food Technologist (R&D), Tasty Bite Eatables
