2.1: Summary of a selection of descriptive methods

Flavour profile method

1. Used to describe and quantify aroma, flavour, and aftertaste components. First published descriptive method.

2. Assessors, typically 5-8, are selected on basis of taste and aroma ability.

3. Assessors are trained on how to assess aroma, flavour, and aftertaste with reference standards.

4. Panel leader can also act as assessor.

5. An eight-point ordinal category scale is generally used to estimate intensity of each attribute e.g. 0 = absent through to 4 = strong.

6. Uses a fixed frame of reference to express intensities (absolute scale).

7. Originally consensus scores for each attribute were developed after discussion but statistical analysis can be used when individual measurements are used.

Texture Profile Method (See ISO 11036:1994. Sensory analysis -- Methodology -- Texture profile)

1. Used to describe and quantify textural and mouthfeel attributes.

2. Developed based on the Flavour Profile Method.

3. Assessors, typically 6-10, are selected on basis of their ability to discriminate known textural differences on the product type of interest.

4. Assessors are trained by panel leader how to assess texture in a standardised way.

5. Panel leader can also act as assessor.

6. Predetermined reference scales for textural attributes are used.

7. Assessors score products for attributes by consensus or individually using a scale (category, continuous line or Magnitude estimation) on which they have been trained.

8. Statistical analysis can be used when individual measurements are used.

QDA

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

2. Developed as an improvement upon the Flavour Profile method (i.e. panel leader is not a participant, data is not generated through consensus but analysable by statistics and unstructured line scales are used).

3. Assessors, typically 10-12, are selected from a large pool of individuals according to their ability to describe and discriminate between products in the category to be ultimately tested.

4. Panel leader acts as a facilitator and does not direct panel discussion nor scores the products. Assessors create the list of attributes as a group.

5. Each assessor's performance is monitored via the use of replications and compared to that of the whole panel.

6. No absolute measures as in the Flavour and Texture Profile Methods (although sometimes used this way).

7. Individual product assessment takes place using continuous line scales.

8. Data are analysed statistically using various statistical methods such as Analysis of Variance (ANOVA) and Principal Component Analysis (PCA) (see section 5.3).

Spectrum method

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

2. Developed based on the Flavour and Texture Profile Methods.

3. Assessors, typically 12-15, are selected from a large pool of individuals according to their ability to describe and discriminate between products in the category to be ultimately tested.

4. Panel leader can also act as assessor.

5. Assessors are trained to assess the flavour and texture of products in a standardised way using a standardised lexicon with no purely product specific terms.

6. Uses category anchors provided by non-associated foods e.g. hardness scale; low = cream cheese to high = boiled sweet.

7. Standard 15-point 'absolute intensity' scales are used.

8. Data are analysed statistically using various statistical methods such as ANOVA and PCA.