《BS-5929-6-1989 ISO-8587-1988.pdf》由会员分享，可在线阅读，更多相关《BS-5929-6-1989 ISO-8587-1988.pdf（14页珍藏版）》请在三一文库上搜索。
1、BRITISH STANDARD BS 5929-6: 1989 ISO 8587:1988 Methods for sensory analysis of food Part 6: Ranking Licensed Copy: sheffieldun sheffieldun, na, Mon Dec 04 15:00:54 GMT+00:00 2006, Uncontrolled Copy, (c) BSI BS 5929-6:1989 This British Standard, having been prepared under the direction of the Food an
2、d Agriculture Standards Policy Committee, was published under the authority of the Board of BSI and comes into effect on 31 August 1989 BSI 06-1999 The following BSI references relate to the work on this standard: Committee reference FAC/20 Draft for comment 85/52045 DC ISBN 0 580 17493 X Committees
3、 responsible for this British Standard The preparation of this British Standard was entrusted by the Food and Agriculture Standards Committee (FAC/-) to Technical Committee FAC/20, upon which the following bodies were represented: AFRC Institute of Food Research Association for Consumer Research (AC
4、RE) British Essence Manufacturers Association British Food Manufacturing Industries Research Association British Psychological Society Campden Food and Drink Research Association Co-operative Union Coffee Trade Federation Ltd. Department of Trade and Industry (Laboratory of the Government Chemist) F
5、lour Milling and Baking Research Association Food and Drink Federation Institute of Food Science and Technology Institute of Masters of Wine Joint Committee of the Milk Marketing Board and the Dairy Trade Federation Ministry of Agriculture, Fisheries and Food Royal Society of Chemistry Society of Ch
6、emical Industry Society of Cosmetic Scientists Society of Dairy Technology University of Reading University of Strathclyde Amendments issued since publication Amd. No.Date of issueComments Licensed Copy: sheffieldun sheffieldun, na, Mon Dec 04 15:00:54 GMT+00:00 2006, Uncontrolled Copy, (c) BSI BS 5
7、929-6:1989 BSI 06-1999i Contents Page Committees responsibleInside front cover National forewordii 1Scope1 2Normative references1 3Definitions1 4Principle1 5Apparatus1 6Sampling1 7General test requirements1 8Preparation of test samples2 9Procedure2 10Expression and interpretation of results3 11Test
8、report6 Annex A (normative) Practical example of application7 Annex B (informative) Specimen answer form8 Table 1 Summary of assessors evaluations3 Table 2 Decoding and calculation of rank sums for the example given in Table 13 Table 3 Approximate critical values from the Friedman test (levels of 0,
9、05 and 0,01)4 Table 4 Critical values of 2 distribution (levels of 0,05 and 0,01)4 Table 5 Critical values from the Page test6 Table A.1 Example of evaluation7 Publications referred toInside back cover Licensed Copy: sheffieldun sheffieldun, na, Mon Dec 04 15:00:54 GMT+00:00 2006, Uncontrolled Copy,
10、 (c) BSI BS 5929-6:1989 ii BSI 06-1999 National foreword This Part of BS 5929 has been prepared under the direction of the Food and Agriculture Standards Policy Committee. It is identical with ISO 8587:1988 “Sensory analysis Methodology Ranking”, published by the International Organization for Stand
11、ardization (ISO), and in the preparation of which the United Kingdom played a full part. The various parts of ISO 5492, referred to in clauses 2 and 3, are being published in a single classified form and it is intended that the resulting international standard will be implemented as an identical (du
12、al-numbered) revision of BS 5098:1975 “Glossary of terms relating to sensory analysis of food”. In the mean time the Technical Committee has reviewed the provisions of ISO 5492 and has decided that they are acceptable for use in conjunction with this Part of BS 5929 for the definitions of any terms
13、not given in BS 5098. A British Standard does not purport to include all the necessary provisions of a contract. Users of British Standards are responsible for their correct application. Compliance with a British Standard does not of itself confer immunity from legal obligations. Cross-references In
14、ternational standardCorresponding British Standard ISO 3534:1977BS 5532 Statistical terminology Part 1:1978 Glossary of terms relating to probability and general terms relating to statistics (Identical) BS 5929 Methods for sensory analysis of food ISO 5495:1983Part 2:1982 Paired comparison test (Tec
15、hnically equivalent) ISO 6658:1985Part 1:1986 General guide to methodology (Identical) ISO 8589:1988aBS 7183:1989 Guide to the design of test rooms for the sensory analysis of food (Identical) a Indicated in clause 2 as “to be published” but now published. Summary of pages This document comprises a
16、front cover, an inside front cover, pages i and ii, pages 1 to 8, an inside back cover and a back cover. This standard has been updated (see copyright date) and may have had amendments incorporated. This will be indicated in the amendment table on the inside front cover. Licensed Copy: sheffieldun s
17、heffieldun, na, Mon Dec 04 15:00:54 GMT+00:00 2006, Uncontrolled Copy, (c) BSI ISO 8587:1988 (E) BSI 06-19991 1 Scope This International Standard describes a method for sensory evaluation of test samples with the aim of placing a series of test samples in a ranking order. The method applies to perfo
18、rming multisample difference testing using the criterion of intensity of single attributes, components of attributes or total impression. It is especially recommended for the pre-sorting of test samples (to be followed by the application of other test methods) or when other methods are beyond the ca
19、pabilities of the assessors to use reliably. Among other things, the method enables the influence of different raw materials, processing, treatment, packaging and storage to be determined. It may also be suitable for use in the training of assessors. 2 Normative references The following standards co
20、ntain provisions which, through reference in this text, constitute provisions of this International Standard. At the time of publication, the editions indicated were valid. All standards are subject to revision, and parties to agreements based on this International Standard are encouraged to investi
21、gate the possibility of applying the most recent editions of the standards listed below. Members of IEC and ISO maintain registers of currently valid International Standards. ISO 5492-1:1977, Sensory analysis Vocabulary Part 1. ISO 5492-2:1978, Sensory analysis Vocabulary Part 2. ISO 5492-3:1979, Se
22、nsory analysis Vocabulary Part 3. ISO 5492-4:1981, Sensory analysis Vocabulary Part 4. ISO 5492-5:1983, Sensory analysis Vocabulary Part 5. ISO 5492-6:1985, Sensory analysis Vocabulary Part 6. ISO 5495:1983, Sensory analysis Methodology Paired comparison test. ISO 6658:1985, Sensory analysis Methodo
23、logy General guidance. ISO 8589:, Sensory analysis General guidance for the design of test rooms1). 3 Definitions See ISO 5492 for the definition of terms relating to sensory analysis. Statistical terms are used as defined in accordance with ISO 3534:1977, Statistics Vocabulary and symbols. 4 Princi
24、ple Simultaneous presentation of several test samples in random order to assessors. Ranking of the samples according to a specified criterion (for example, total impression, particular attribute or specific characteristic of an attribute). If a reference sample is used, it is placed unidentified amo
25、ng the other samples. Statistical evaluation of the test results. 5 Apparatus The apparatus shall be selected by the test supervisor, according to the nature of the product, the number of samples, etc., and shall in no way affect the test results. If standardized apparatus corresponds to the needs o
26、f the test, it shall be used. 6 Sampling Refer to the standards relating to sampling, for sensory analysis, of the product or products to be examined. The method of sampling shall take account of the test objective and, if there is no standard for the product concerned, shall be agreed between the i
27、nterested parties. 7 General test requirements 7.1 Room For the design of, and the conditions in, the room in which tests are to be conducted, see ISO 8589. 7.2 Assessors 7.2.1 Qualification For the conditions which the assessors shall fulfil, see ISO 6658. All the assessors shall preferably have th
28、e same level of qualification, this level being chosen according to the purpose of the test (for example, untrained assessors, selected assessors or experts). 1) To be published. Licensed Copy: sheffieldun sheffieldun, na, Mon Dec 04 15:00:54 GMT+00:00 2006, Uncontrolled Copy, (c) BSI ISO 8587:1988
29、(E) 2 BSI 06-1999 7.2.2 Number of assessors The number of assessors depends on the aim of the test. In general at least five selected assessors are necessary. For statistical analysis of the results, the larger the number of assessors, the greater the probability of revealing differences in rank bet
30、ween products. All assessors shall work under the same test conditions. 7.3 Preliminary discussion The assessors shall be informed of the purpose of the test. If necessary, a demonstration can be arranged on the basis of ranking. It is essential in this test to ensure common understanding of the cri
31、terion tested by all assessors. The preliminary discussion shall not influence the test results. 8 Preparation of test samples Where applicable, specify in accordance with the purpose of the test the kind of preparation and the presentation of the sample; the number of samples; the temperature of sa
32、mples (identical for all samples in the same test); any masking of single attributes (for example the use of coloured lamps to remove the effects of colour; homogenizing of samples to remove the effects of texture). For each sample, the quantity of product presented to the assessor shall be the same
33、 and shall be sufficient to allow an evaluation of each of the samples as many times as required. The number of samples to be ranked will be determined in accordance with the degree of difficulty of the test. (For example, for assessing samples of intense flavour, the number shall be kept small; on
34、the other hand, for ranking according to the criterion “colour”, the number of samples may be greater.) The vessels containing the test samples shall be coded, preferably using three-digit random numbers. The coding shall be different for each test. 9 Procedure 9.1 Presentation of samples The assess
35、ors shall not be able to draw conclusions as to the nature of the samples from the way in which they are presented. Therefore, the various samples of a series shall be presented in an identical fashion (same apparatus, same vessels and same quantities of products). 9.2 Reference samples Reference sa
36、mples may be introduced. In this case, these samples are introduced unidentified into the series of samples. 9.3 Test technique The assessors evaluate P samples presented in random order and place them in a certain rank order, restricting themselves to the specified criterion. The same series of sam
37、ples may be presented to each assessor one or more times with different codes. As instructed by the test supervisor, the assessors assign the rank 1 to the sample with the strongest or weakest intensity of the attribute to be tested (for example, hardest/softest, sweetest/least sweet). The ranks 2 t
38、o P are given to the other samples in sequence. The assessors should be instructed to avoid tied rankings. They should be told that, even when they cannot find much difference between two samples, they should make a best guess. However, if they are unable to differentiate between samples, they shall
39、 make a note of this on the answer form (see 9.4) in the section reserved for comments. It is recommended that assessors set up a provisional rank order first and then verify it by further evaluation into order of increasing intensity. Only one attribute may be considered. If a rank order for severa
40、l attributes is required, the test supervisor shall set up a separate experiment for each attribute. Instructions specific to certain products shall be indicated (for example, “stir before assessing the odour”). In the case of gustatory stimuli, the assessors may be invited to use auxiliary products
41、, such as water, weak tea or white bread, to neutralize sensations between assessments. 9.4 Answer form The ranks of the individual samples shall be recorded on an answer form. A specimen answer form is given in Annex B. Depending on the purpose of the test and the test samples, it may be necessary
42、to record further information. Licensed Copy: sheffieldun sheffieldun, na, Mon Dec 04 15:00:54 GMT+00:00 2006, Uncontrolled Copy, (c) BSI ISO 8587:1988 (E) BSI 06-19993 10 Expression and interpretation of results 10.1 Collation of results If necessary, summarize in a table the evaluations recorded o
43、n the answer forms by each of the assessors, for each test and for each attribute, indicating tied ranking samples by an “equal to” sign (see Table 1). Table 1 Summary of assessors evaluations NOTEFor convenience, the letters A, B, C and D have been used in Table 1 and Table 2. In Table 1, the lette
44、rs represent the coded samples using random three-figure numbers as presented to assessors. In Table 2, the letters represent the actual names of the samples. 10.2 Decoding of the samples and calculation of the rank sums Decode the samples and tabulate the rank orders given to each sample by each as
45、sessor. Where there are tied rankings, note the mean rank (see Table 2). Calculate the rank sum for each sample on the basis of the results by adding column sums in Table 2. By comparing the rank sums for the samples it is possible to obtain an evaluation of the differences between the samples. Tabl
46、e 2 Decoding and calculation of rank sums for the example given in Table 1 10.3 Statistical interpretation From the numerous tests featured in the literature, the following are recommended: the Friedman test2), which, in a very general way, gives the maximum opportunities for demonstrating recogniti
47、on by the assessors of differences between the samples3); and the Page test4), where there is a pre-determined order of the samples. 10.3.1 Friedman test 10.3.1.1 Overall comparison of all samples Calculate the Friedman value F, as follows: where J is the number of assessors; P is the number of samp
48、les (or products); R1, R2 . . ., RP are the rank sums attributed to the P samples for the J assessors. Then compare F with the critical values from Table 3. Assessor Samples Rank sums ABCD 1123410 241,51,5310 3133310 4134210 5312410 Rank sums for the samples 1010,513,51650 NOTEThe row totals are ide
49、ntical and equal to 0,5 P (P + 1). 2) FRIEDMAN, M. The use of ranks to avoid the assumptions of normality implicit in the analysis of variance. Journal of the American Statistical Association, 32 (1937) pp. 675-701. 3) In response to the question, “Are there differences in assessment between the samples?”, other techniques are also available: the technique proposed by Fisher, which consists simply of carrying out an analysis of variance after first converting the ranks into scores; two other techniques w