Table 2.2: summary of alternative descriptive methods

Free choice profiling

1. Can be used to describe and quantify all sensory aspects of a product or specific modalities may be assessed.

2. Developed as a 'rapid' descriptive method.

3. Assessors, a minimum of 10, are screened for their sensory abilities and use their own individual attributes and scales when evaluating the products. The numbers of attributes used can vary for each assessor.

4. There is no panel leader as such as no discussions or agreement about attributes are required - the panel leader role is more of an organiser.

5. There is very little panel training.

6. Assessors use their own choice of scale, but must use it consistently, to assess the intensities of the attributes they have generated one sample at a time.

7. No training on intensities is required.

8. Generalized Procrustes Analysis (GPA) is used to analyse the results to generate a 'consensus space' in which each individual's data are plotted. (Delegates only need demonstrate an awareness of GPA and do not need to be able to perform it.)

Flash Profiling

1. Can be used to describe and rank the intensities of all sensory aspects of a product or specific attributes/modalities may be assessed.

2. Developed by Sieffermann (2000) as a 'rapid' descriptive method and is a combination of FCP and ranking methods.

3. Can use screened, trained, and experienced assessors (minimum 10), consumers (40-50) or product experts, who generate their own attributes and rank the samples accordingly.

4. A quick method for determining the relative sensory positioning of a set of products. Useful tool for rapid market analyses.

5. There is very little training required.

6. All the samples are presented at the same time allowing the assessors to make direct comparison for description and ranking the samples for each attribute they choose. This can make the choice of flash profiling difficult for some product types (e.g. temperature sensitive products such as ice creams or strongly flavoured products due to carryover).

7. No training on intensities is required.

8. Data is analysed using GPA to assess the consensus between assessors.

Check-all-that-apply (CATA) (Qualitative sensory profile ISO 13299:2016(E))

1. Can be used to describe but not quantify (unless the extension to rate all-that-apply (RATA ) is employed) all sensory aspects of a product or specific modalities may be assessed.

2. Developed as a 'rapid' descriptive method, however hedonic and emotion questions can also be used to help aid interpretation of liking and description.

3. Assessors (16 to 30) or consumers (50-100) are presented with a list of terms to 'check all that apply'. Terms can be chosen by the development team or most often by prior consumer studies (focus groups, open-ended questions in questionnaires).

4. There is no panel leader as such as no discussions or agreement about attributes are required - the panel leader role is more of an organiser.

5. There is very little panel training.

6. Samples are generally assessed one at a time.

7. No training on intensities is required as the data generated is frequencies (not rankings or intensities).

Correspondence Analysis or Multiple Factor Analysis applied to frequency matrix to produce a sensory map.

Projective mapping and sorting

1. Can be used to compare samples across all sensory aspects of a product or specific modalities based on perception of overall sample set differences

2. Variations in the methodology include:

Sorting, in which samples are sorted into categories. Variants include free sorting and hierarchical sorting.

Projective mapping, which was developed as a 'rapid' method in 1994 and measures the sensory distance perceived between products. Napping® is a commonly used variant (the word comes from "nappe" meaning tablecloth in French) which involves positioning products on a piece of blank paper or computer screen.

Ultra flash profiling is another variant that enables product description.

3. Sorting can be carried out with consumers or by trained assessors, but if the idea is to assess products holistically, consumers are recommended. Numbers of assessors varies in the literature.

4. There is no panel leader as such as no discussions or agreement about attributes are required - the panel leader role is more of an organiser.

5. There is very little panel training required.

6. All the samples are presented at the same time, generally in one session, allowing the assessors to make direct comparisons between the samples. This can make the choice of these methods difficult for some product types (e.g. temperature sensitive products such as ice creams or strongly flavoured products due to carryover).

7. No training on intensities is required.

Multiple Factor Analysis (MFA) is performed on projective mapping data because each respondent has two variables associated: the X and the Y coordinates of the products on their map. Sorting data, in the form of a similarity matrix, is analysed using multi-dimensional scaling (MDS).