Organoleptic (sensory) measurement of spoilage

Quality assessment methods

The food manufacturer has to produce a product at a quality which will satisfy:

(a) the customer
and
(b) statutory food legislation.

Much has been written on the analytical techniques developed to establish the nutritional and chemical composition of foods to ensure their compliance with food law, and it is not proposed to explore this area in this lecture.

In many ways, the quality requirements of the customer are more difficult to satisfy since the customer assesses the quality of a product entirely by subjective means, i.e. by sensory evaluation of a food's appearance, colour, odour, taste and texture, plus the visual appeal of its packaging and presentation.

Hence, it is the responsibility of the manufacturer to develop methods which can, as accurately as possible, evaluate the sensory properties of a food which the customer finds important. This will often mean using subjective methods of testing such as sensory (taste panel) methods or, in some cases, it may be possible to use objective methods.

However, it must be remembered that objective measurements of food quality are preferable only if the objective tests can provide a precise measure of the subjective quality being considered.

Sensory methods

A sensory method is one in which subjective measurements are made by individuals. Numerical scoring systems may be developed for such methods, or methods of ranking, or results may be expressed simply in attribute form, e.g. pass/fail, acceptable/defective.

Objective methods

An objective method uses a calibrated scientific instrument to measure a specific quality parameter usually on a numerical scale.

Problems of assessing fish quality

1. Many hundreds of species are sold throughout the world, all with distinct physical, chemical and sensory characteristics.

2. Fish, alone among the major items of food, are susceptible to virtually no control before harvesting or slaughtering. In addition, they are one of the most perishable items of food eaten.

3. Fish freshness/spoilage may be investigated objectively by assessment based on, for example:

(a) Physical changes, e.g., measurement of conductivity using the Torrymeter.

(b) Chemical and biochemical tests, e.g.

- tests dependent on bacterial action such as estimation of trimethylamine (TMA);
- tests dependent on autolytic action such as enzymic assays of nucleotide breakdown products, e.g., hypoxanthine;
- tests dependent on fat oxidation, such as peroxide value estimation.

(c) Bacteriological changes, e.g., counting and identification of various micro organisms.

However, none of these methods produces results which may be regarded as providing a precise measure of consumer acceptability. Therefore, any system adopted to measure the quality of fish must, at some stage, be related to consumer acceptance. This must involve sensory ('taste') panels and consumer research. in the following sections, it is planned to consider the problems and methods of establishing a taste panel and assessment of results.

Classification of sensory characteristics

Sensory appreciation of food quality may be divided into the following categories:

1. Appearance and colour;
2. Odour and taste;
3. Texture.

The above three categories cannot be treated in isolation since sensory evaluation is often a combination of several overlapping factors, e.g., 'flavour' includes elements of odour, taste, texture and even the psychological effect of colour.

Taste panels

Aims

Taste panel techniques may be used in many different ways:

1. To characterise sensory changes in foodstuffs, e.g., changes that occur during spoilage or changes brought about by processing methods such as freezing or dehydration.

2. To distinguish (often with the hope of failure) between batches or samples of a particular foodstuff.

3. To ascertain whether 'qualify' can be represented by a simple numerical index or whether it is multidimensional.

4. To help establish standards, in some defined area, of raw or unprocessed food products.

5. To grade products according to some agreed quality classification system.

6. To help establish a usable relation between sensory data and consumer acceptability.

Methods to achieve the above aims roughly divide into two categories:

(a) Product rating
Grading on some agreed numerical or descriptive/hedonic scale is carried out by a small panel of well-trained judges.

(b) Difference testing
Usually done by a panel of sensitive judges, without any great expertise or elaborate training.

Panel selection and training

Prospective taste panel members must be in sound health, self-motivated and of even temperament. They should first be screened for primary taste sensitivity and reliability. Members should be consistently able to detect and recognise the following levels of the four primary taste components:

Sweetness	- 1.5 per cent sucrose
Saltiness	- 0.15 per cent sodium chloride
Sourness	- 0.06 per cent citric acid
Bitterness	- 0.005 per cent quinine sulphate

Many individuals will be able to detect levels considerably lower than these values. It is also important to remember that initial detection of a component (say differentiation from water) is likely to occur at a lower concentration than recognition of the primary taste concerned.

If evaluation of odour is a major part of the sensory test then prospective judges must also be able to demonstrate a good 'odour memory'. On a general level, it is wise to find out how well the candidates will perform on an odour recognition test carried out with stoppered bottles of some of the commoner chemical smells such as acetic acid, ammonia, amyl acetate, aniseed, benzaldehyde, linseed oil, menthol, peppermint, vanillin etc.

Finally, prospective members should be able to display discriminatory skill with different qualities of the foodstuff under test. Discriminatory skill need not be general, e.g. a good wine taster may not be a good judge of teas or fish.

Having chosen members for the 'taste panel', a preliminary training period should be carried out, designed to acquaint the tasters with the quality factors involved in the fish product under test. The group should be asked to evaluate a selection of fish of the same type, at different stages in deterioration, from fresh to wholly unacceptable.

Tasting room

Control of environmental factors is universally recognised as being of value in sensory work with foods. Thus, a special 'taste panel' room where as many variables as possible can be controlled is highly recommended.

For maximum privacy and concentration, the use of screened tasting booths is recommended.
Although natural or white fluorescent lighting should be used where possible, a coloured light facility should be made available to disguise any minor colour differences between samples that may occur on occasion and which would otherwise distract the taster.

All possible sources of extraneous odour should be excluded, e.g. samples should be prepared in a separate room to eliminate cooking odours. Beware also of odours from detergents, floor polishes and other cleaning compounds.

Preparation of samples

The cooking method selected should not add any extraneous flavour to the food, e.g. frying is a poor method, but all sample flavour should be retained where possible, e.g. for fish, steaming "casserole) is recommended.

All samples should be presented at the same temperature.

In 'difference testing', the difficulty of presenting a 'standard reference sample' is acknowledged since: (a) any standard will deteriorate with time and (b) any standard will be quickly used.

Many workers use a standard control taken from a frozen stock of acceptable material.

Bias

In order to neutralise errors associated with the order of presentation of samples, the order in which items are presented should be balanced between the possible alternatives, e.g. in a triangular test:

AAB	ABA	BAA
ABB	BAB	BBA

Samples should also be labelled in such a way as to eliminate any subconscious bias in the taster. For example, labelling samples with

Fatigue

The maximum number of samples which can be reasonably assessed in one session will vary depending on the nature of the product. In general, fewer strongly tasting foods may be tested in one session than more bland products.

In order to avoid transfer ('carry-over') of flavour from one sample to the next, a warm water rinse between samples is recommended. (Some authors also recommend the use of white bread, unsalted biscuits, apple, lemon juice etc.).

Some commonly used test designs
(a) Methods of product rating

The sensory panel should consist of 4 to 6 individuals, who have undergone a period of intensive training both with the product under test and with the procedures being used.

Ranking

Ranking tests require that judges arrange a series of two or more samples in an ascending or descending order of intensity of a specific characteristic. Samples may be ranked in order of degree of acceptability, or in order of general quality, or by specific attributes of colour, texture or flavour intensity.

Grading

The panel is asked to grade samples according to some agreed numerical or descriptive hedonic scale.

Numerical scoring systems: Freshness, or degree of spoilage, may be assessed in a raw fish from:

(i) the general appearance of the fish including that of the eyes, surface slime, and texture of the flesh and
(ii) the odour of the gills and belly cavity.

In addition, marks may also be awarded for cooked fish odour, texture and flavour.

Various numerical scoring systems have been devised to cover some or all of the above sensory parameters. The most intensive investigation of this type of sensory evaluation has been that of Shewan and co-workers, who have devised detailed descriptive schemes in which numerical scores are given to:

(i) Raw fish - appearance, odour and texture;
(ii) Cooked fish - flavour, odour and texture.

Such schemes have been devised for cod, haddock, whiting and redfish, and similar schemes have been developed for various other species.

Generally speaking, the full Shewan's scheme is probably too complex for regular commercial use although abbreviated versions may prove useful for routine quality grading.

Hedonic seales: For example,

Excellent	Good	Average	Poor	Bad
Like	Like	Neither like	Dislike	Dislike
definitely	slightly	nor dislike	slightly	definitely

For a scoring or grading system to be effective certain prerequisites must be met:

1. With numerical scoring systems, quality factors must be properly 'weighted' to reflect their relative importance.

2. The scale should be such that a difference in score reflects a reproducible variation in the quality factor being scored, i.e. scale not too large with too many elements.

3. Agreement is necessary between judges as to quality relating to specific scores given.

4. Whenever possible, physical and chemical analysis of the commodity should be carried out to supplement sensory evaluation.

(b) Methods of difference testing

The panel should consist of 10 to 20 individuals with, ideally, 3 or 4 replications per judge per 'difference test'.

The panel do not require any intensive training as long as members are reasonably motivated and have passed the sensitivity requirements.

The three most commonly used test designs are the pair comparison, duo-trio test and the triangular test.

Pair comparison involves simultaneous presentation of one coded sample each of material A and B with the question: -

'Which is the...................er of the two?' or
'Which is the regular sample?'

In the Duo-trio test, three samples, two of A with one of B, or one of A with two of B, are presented. One of the duplicates is coded, say, S and the other samples are coded, say, 1 and 2 and the question asked is 'either 1 or 2 is identical with S; which is it?'

In the Triangular test, three coded samples are presented, of which two are identical and one is different but, this time, the question is 'which one is the odd sample?'

Statistical calculations can be carried out to establish whether the number of 'correct' answers obtained is sufficient to demonstrate a significant difference in the flavour of the 'odd' sample.

Conclusions

The introduction of sensory panel evaluation of fish quality requires research, good organisation and proper training.

An established visual and organoleptic scale is probably one of the easiest and least expensive ways of evaluating fish spoilage. It has the advantage of meaning something to the fisherman, the fish seller and the consumer. Its main disadvantage is that it is often open to discussion and disagreement.

Each country should develop its own standards under its own conditions. Consumer acceptability must always be the criterion on which to base methods. If fish products are for export, retraining of taste panel personnel and review of quality methods may be necessary to meet new market requirements.

References

1. AMERINE, M. A., PANGBORN, R. M. and ROESSLER, E. B. (1965) Principles of sensory evaluation of food. New York and London: Academic Press, 602 pp.

2. EHRENBERG, A. S. C. and SHEWAN, J. M. (1953) The objective approach to sensory tests of food. Journal of the Science of Food and Agriculture, 4, 482 - 490.

3. EHRENBERG, A. S. C. and SHEWAN, J. M. (1959) Applied Statistics, 8, 186.

4. HERSCHDOERFER, S. M. (Ed.) (1967 - 72) Quality control in the food industry, Vols. 1 and 2, London: Academic Press.

5. KREUZER, R. (Ed.) (1971) Fish inspection and quality control. Published by arrangement with the Food and Agriculture Organization of the United Nations. West Byfleet, Surrey: Fishing News (Books) Ltd.

6. SHEWAN, J. M., MACINTOSH, Ruth G., TUCKER, C. G. and EHRENBERG, A. S. C. (1953) The development of a numerical scoring system for the sensory assessment of the spoilage of wet white fish stored in ice. Journal of the Science of Food and Agriculture, 4, 283 - 297.