Sensory evaluation of foods.pdf
Transcript
Types of sensory tests BTF232TC2 Sensory evaluation of foods Dr. Subajiny Sivakanthan Senior Lecturer Faculty of Agriculture University of Jaffna 1 19/09/2024 ...
Types of sensory tests BTF232TC2 Sensory evaluation of foods Dr. Subajiny Sivakanthan Senior Lecturer Faculty of Agriculture University of Jaffna 1 19/09/2024 Reference book Meilgaard, M.C., Civille, G. V. and Carr, B. T. 2016. Sensory Evaluation Techniques, 5th edition, CRC Press. 2 19/09/2024 Types of sensory tests Discriminative test Paired – comparison Duo – trio Triangle Ranking Rating difference / scalar difference from control Sensitivity tests Threshold test Analytical tests Dilution test Descriptive tests Attribute rating Category scaling Ratio scaling Descriptive analysis Flavor profile analysis Texture profile Quantitative descriptive analysis Paired preferences Ranking Affective tests Rating Hedonic scale 3 Food action scale Types of sensory tests Analytical tests (Product oriented Affective tests (Consumer oriented test) test) Evaluate differences or similarities, Evaluate preference and or acceptance and quality and or quantity of sensory or opinion of a product characters of a product Panelist randomly selected, untrained Panelists screened for interests, ability to consumers of test product discriminate differences and reproduce results, trained to function as a human No recommended number analytical instrument 50 -100 panelist usually considered No recommended number for number adequate of panelists. 10 – often used, 5 – 30 – 50 panelist considered adequate for recommended minimum in-house evaluation 4 Analytical tests Descriptive test Discriminative test describing characteristics of a determining if differences exist between two or more products product and/or measuring any differences between products Is product A identical to product B? What does this product taste like? Find the two similar products What are the three most among the three samples provided (Eg; AAB or ABB) important texture attributes you perceive in this product? Find the odd sample among the three samples provided (Eg; For which sensory attributes are AAB or ABB) the differences between product A and B most marked? Analytical tests 1. Discriminative 2. Descriptive Difference (measures simply (Measurers qualitative and whether samples are different) quantitative characteristic) Paired – comparison Attribute rating Duo – trio Category scaling Triangle ratio scaling Ranking Descriptive analysis Rating difference / scalar flavor profile analysis difference from control texture profile Sensitivity test quantitative descriptive analysis Threshold Dilution 6 19/09/2024 Affective tests Paired preferences Ranking Rating Hedonic scale Food action scale 7 19/09/2024 1. Paired comparison test Two coded samples are evaluated simultaneously or sequencially Simple difference (same or different test) To determine whether a sensory difference exists between two products It is effective to determine whether product difference result from a change in ingredients, processing, packaging or storage 8 The panelist is told beforehand that the samples in each trial set to be tested may be identical or different. For half of the panels, present same sample twice (Placebo) (AA or BB) and for half of the panels present different samples (AB or BA) Analyze results by comparing number of “different” responses for the same samples and “different” responses for the different samples using chi- square test Example Problem/Situation: A fruit drink manufacturer received a report from market that drink A has less flavour. Thus manufacturer developed drink B using high flavour. Objective: To compare drink A and B to determine whether a small but significant increase in flavour has been attained. Test design: Paired comparison/simple difference 10 19/09/2024 SCORE SHEET Simple Difference test Name: Date: Type of sample: Instructions: Taste pair from left to right Determine if samples are same or different. Mark your response below Products are same ……… Products are different ……… Comments: 11 SCORE SHEET Simple Difference test (Multiple paired comparisons) Name: Date: Type of sample: Characteristic studied: Instructions: Taste each pair from left to right and determine if samples within each pair are same or different and enter your verdict below. Test pairs Same/different ……… ……… ……… ……… ……… ……… ……… ……… ……… Comments: 12 Example Data Subjects received Matched pair Unmatched pair Total (AA or BB) (AB or BA) Subjects said Same 17 9 26 Different 13 21 34 Total 30 30 60 13 19/09/2024 Directional difference The panelist is asked to choose the sample within each pair that has the greatest amount of given character (Eg. High flavor, high sweetness) A forced choice is required (panelist has to choose one sample even both samples looks similar). Number of sample per test – two Prepare equal number of combinations AB and BA Ask the subject to taste the product left to right and fill in the score sheet 14 Example Problem/Situation: A fruit drink manufacturer received a report from market that drink A has less flavour. Thus manufacturer developed drink B using high flavour. Objective: To compare drink A and B to determine whether a small but significant increase in flavour has been attained. Test design: Paired comparison/directional difference 15 19/09/2024 SCORE SHEET Directional Difference test Name: Date: Type of sample: Fruit Drink Characteristic studied: Flavour Instructions: Instructions: Taste samples from left to right and determine which sample contains high flavor. If no difference is apparent, enter your best guess, however uncertain. Samples contains high flavour ……… ……… Comments: 16 SCORE SHEET Directional Difference test (Multiple paired comparisons) Name: Date: Type of sample: table syrup Characteristic studied: mouthfeel (thickness) Instructions: Taste first sample pair from left to right and note which sample is thicker. Indicate by placing x next to the code. If you perceive no difference, please make a best guess. Continue until all 3 pairs have been evaluated. Rinse with water as needed to clear your palate. Pair no. Left sample Right sample 1 ……… ……… 2 ……… ……… 3 ……… ……… Comments: 17 Analysis and interpret the results Samples are coded as 812 (Sample B) and 691 (Sample A). Number of panelists: 40 Sample 812 is chosen by 28 to contain high flavour From the table 19.12 (n=40, α=0.05), it can be concluded that there is a significant difference because the value (28) is higher than table value (27). Interpret Suggest that sample A is used in future as it has significantly high flavour 18 19/09/2024 2. Duo – trio test Employs three samples – Two identical, One different (AAB or ABB) One sample is identified as the reference and presented first Next the two coded samples are presented (one of which matches the reference sample) One of these two samples is identical to the standard / control Count the number of correct replies and refer to the statistical table (19.10) for interpretation. Do not count ‘no difference’, panelists must guess if in doubt Used to determine whether product differences result from a change in ingredients, processing, packaging or storage Example Problem/Situation: A fruit drink manufacturer received a report from market that drink A has less colour. Thus manufacturer developed drink B to have high colour. Objective: To compare drink A and B to determine whether a significant increase in colour has been attained. Test design: Duo-trio test 20 19/09/2024 SCORE SHEET Duo-trio test Name: Date: Type of sample: Fruit drink Characteristic studied: Colour Instructions: Observe the sample from left to right. The left-hand sample is a reference. Determine which of the two samples matches the reference in colour and indicate by placing an x. If no difference is apparent between two samples, you must guess. Reference Code Code Comments: 21 Analysis and interpretation Example: If only 21 out of 40 subjects chose the correct match to the designated reference, according to the Table 19.10, twenty six correct responses are required (at alpha 5%). Thus, the duo-trio test failed to detect any significant difference in colour of two samples See example 7.5 22 19/09/2024 3. Triangle test Used to determine whether a sensory difference exists between two products. Employ three coded sample – Two identical, One different (AAB or ABB) Presented simultaneously None of the sample is identified on the reference/control Reference and experimental treatment (samples) – systematically varied The panelist has to determine which of the three sample differs from the other two (odd sample) 23Count the number of correct replies and refer table 19.8 19/09/2024 Score sheet Triangle test Name: Date: Type of sample: Instructions: Taste the samples from left to right. Two are identical. Determine which the odd sample is. Which is the odd sample? Comments …….. ………………… 24 19/09/2024 Score sheet Triangle test (Multiple samples) Name: Date: Type of sample: Instructions: Taste the samples from left to right. Two are identical. Determine which the odd sample is. Sets of three samples Which is the Comments odd sample? ………. ………. ………. …….. ………………… ………. ………. ………. ………. ………………… ………. ………. ………. ………. ………………… 25 19/09/2024 4. Simple ranking test Can be done for both consumer oriented and product oriented test Make simultaneous comparisons of several samples on the basis of single characteristic (Eg- Sweetness, Freshness, etc.) No measure of the degree of difference (only ranking) Use to obtain preliminary information on product differences Screen panelists for their ability to discriminate among samples with known differences A control need not be identified 26 19/09/2024 All the test samples are coded Samples are presented to each subject in a balanced and random order Samples can be presented simultaneously or sequentially Subjects are asked to rank according to intensity of characteristic designated No ties are allowed Number of samples per test 2 to 7 Rank totals are obtained for each sample and analyzed statistically with Friedman’s test. 27 19/09/2024 28 19/09/2024 Example (8.5) A manufacturer wishes to compare four artificial sweeteners (A, B, C, and D) for their persistence of sweetness. Objective: To test whether there is a difference among the sweeteners in the persistence of sweetness in the mouth after swallowing. Test design: Ranking test Analysis: Friedman’s test 29 19/09/2024 Pairwise ranking test (difference test) To compare several samples in all possible pairs Friedman’s analysis is used Present to each subject one pair at a time in random order with question (Eg; Which sample is sweeter) Continue until each subject complete all possible pairs If there are four sampes A, B, C and D AB, AC, AD, BC, BD, CD 30 19/09/2024 See example 8.4. Problem: A manufacturer producing blended table syrups wishes to market a product with low thickness. Four corn syrup blends A, B, C and D have been prepared for analysis. Objective: to rank the four blends based on mouthfeel thickness Test design: Pairwise ranking test with Friedman test analysis. Possible pairs, AB, AC, AD, BC, BD and CD can be presented to 12 panelists. 31 19/09/2024 32 19/09/2024 5. Rating difference / scalar difference from control This test is used to determine whether or not a perceptible overall difference exists between one or more samples and a control sample Also give an indication as to the size of any difference perceived. A difference from control test is a useful test to use when other difference tests, such as triangle or duo-trio are not suitable because of the normal heterogeneity of the products to be tested. 33 19/09/2024 The order of presentation of the test and blind control samples should be balanced. Scales or rating used are Line scale Hedonic scale (eg. 9-point hedonic category rating) Category scale Facial hedonic scale often used for children 34 19/09/2024 35 19/09/2024 36 19/09/2024 Eachpanelist is presented with an identified control sample plus one or more test samples. The panelists are asked to rate the size of difference between each test sample and the identified control sample. 37 19/09/2024 Panelists are informed that some of the test samples may be the same as the control sample. The mean difference from control for each test sample is compared with the mean difference from control obtained from the blind presentation of the control sample. The blind control sample is included as a measure of the placebo effect 38 19/09/2024 Score sheet for difference from control test 39 19/09/2024 Dilution Test This test is used to determine the minimum concentration of a substance at which it can be detected. In this test, a series of dilutions are made, and the panelists are asked to identify the lowest concentration where the flavor, taste, or aroma is still perceptible. 40 19/09/2024 Descriptive tests These tests involve trained panelists who provide detailed descriptions of the sensory characteristics of a product, including flavor, aroma, texture, and appearance. The goal is to identify and quantify specific attributes. Attribute Rating: In attribute rating, panelists assess specific sensory attributes of a product, such as sweetness or crispness, on a numerical or graphical scale. The results help quantify the intensity of each attribute. 41 19/09/2024 Category Scaling: This method asks panelists to assign sensory characteristics to predefined categories. For example, the sweetness of a product might be rated as "low," "medium," or "high.“ This method is used to categorize perceptions of intensity. Ratio Scaling: Panelists compare the intensity of one attribute to another and rate them in proportional terms. For example, if one sample is perceived as twice as salty as another, it would be rated accordingly. This approach allows for the evaluation of relative intensity. 42 19/09/2024 Descriptive Analysis: A comprehensive sensory test where trained panelists evaluate and describe the specific sensory characteristics of a product. This involves identifying key attributes and rating their intensity. It is commonly used in product development and quality control. Flavor Profile Analysis: This method focuses specifically on the flavor attributes of a product. Panelists describe the overall flavor experience, including the aroma, taste, aftertaste, and any specific flavor notes. It is used to create a detailed flavor profile. 43 19/09/2024 Texture Profile: This analysis focuses on the textural properties of food, such as hardness, cohesiveness, and chewiness. Panelists evaluate the product’s texture during different stages of mastication (e.g., initial bite, chew-down). It is commonly used for foods where texture is a key quality factor, such as baked goods. Quantitative Descriptive Analysis (QDA): A more advanced version of descriptive analysis, QDA involves a trained panel that quantitatively assesses the intensity of multiple sensory attributes of a product using a structured scale. The results are often visualized in a spider or radar plot to compare products. 44 19/09/2024 “A”–Not A” test This method of discrimination testing comprised of at least two samples; one sample is a previously identified sample (“A”) and one is a test sample. Objective- to determine whether a sensory difference exists between two products, when duo-trio or triangle test cannot be used. Usually used for samples that are very complex stimuli and are mentally confusing to the panellist All samples are presented blindly, and the assessor’s task is to assign the label “A” or “not-A” to each of the samples. 45 19/09/2024 Affective tests Paired preference test Panelists are asked which of the two coded samples they prefer based on the specific attribute. Reasons for preference may be included if desired. Panelists are instructed to choose one, even if both samples seem similar 46 19/09/2024 The two coded samples are presented simultaneously in identical samples container Each order (AB/BA) should be presented an equal number of times Results are analyzed using binomial test The number of judges preferring each sample is totaled and the totals tested for significance 47 19/09/2024 48 19/09/2024 Ranking test Panelists are asked to rank coded samples for acceptance in order from the least acceptable to the most acceptable based on a specific attribute Ties are not allowed Three or more coded samples are presented in identical samples containers 49 19/09/2024 All the samples are presented simultaneously in a balanced random order Data analysis: the ranks assigned to each samples are totaled The samples are then tested for significance difference by comparing the rank totals between or possible pairs of samples using the Friedman test 50 19/09/2024 The differences between all possible rank total pairs are compared with the tabulated critical value based on a specific significant level (5%) and number of panelists and samples involved in the test If the difference between pairs of rank totals is larger than the tabulated critical value, the pair of samples are significantly different at the chosen significant level 51 19/09/2024 Rating test 1.Hedonic test Measure the level /degree of liking (or disliking) for a product or products by a population Category scales –like extremely through neither like nor dislike to dislike extremely 3 point, 5 point, 7 point or 9 point hedonic scales are used. 52 19/09/2024 9-Point Hedonic Scale 5-Point Hedonic Scale 9 = Like Extremely 1 = dislike very much, 8 = Like Very Much 7 = Like Moderately 2 = dislike slightly, 6 = Like Slightly 3 = neither like nor dislike, 5 = Neither Like nor Dislike 4 = like slightly, 4 = Dislike Slightly 5 = like very much 3 = Dislike Moderately 2 = Dislike Very Much 1 = Dislike Extremely 53 19/09/2024 Panelists indicate their degree of liking for each sample by choosing the appropriate category More than one sample may fall within the same category Presentation of samples The coded samples are presented in identical samples containers 54 19/09/2024 The samples order can be randomized for each panelist (or balanced if possible ) Samples may be presented simultaneously (preferred ) or sequentially Data analysis The categories are converted to numerical scores ranging from 1-9 or 1-7, or 1-5 or 1-3 E.g. 1-represents dislike extremely 9-represents like extremely 55 19/09/2024 The numerical scores for each sample are tabulated and analyzed by analysis of variance (ANOVA) This is to find out whether significant difference in mean degree of liking scores exist between or among the samples 56 19/09/2024 57 58 19/09/2024 2. Food action rating scale Nine successive ratio categories ranging from “I would eat “(buy ,use ,etc..) to “I would eat only if I were forced to “ are represented One or more samples may be tested (hedonic and food action rating scale) 59 19/09/2024 60 19/09/2024 Descriptive analysis 61 19/09/2024
