Defra - Department for Environment, Food and Rural Affairs.

Science Search

Science and Research Projects

Return to Science Search homepage   Return to Project List

Collaborative trials as a means of assessment of quality in multivariate analyses - FS2912

Food analysis is often carried out in pursuit of food law enforcement and can result in criminal proceedings for alleged offences against the requirements of the Food Safety Act 1990. For this reason it is essential that analytical quality that is directly related to enforcement of food legislation. The major objective of this project is to determine what are the important factors that influence analytical results obtained from the use of multivariate techniques such as principal components analysis and discriminant functions. This will be achieved by setting up a model system consisting of a calibration set and validation set each containing samples with known values within a fixed population. These samples will be analysed in a co-operative trial (by measurement of UV/VIS spectral data), to test ability to make quantitative analysis of samples in the validation set. The effects of database size, choice of analytical parameter, choice of algorithm and precision of test measurements on trueness, repeatability and reporducibility will be assessed. A model system is used so that the study can be kept under close control with respect to variability and so that trueness can be measured. It also allows analytical costs to be kept to a minimum. Collaborative trials are frequently used to validate analytical methods and a harmonised standard protocol exists for doing this. However, multivariate methods are being used increasingly in food analysis but the question of how to validate such procedures has not yet been addressed.
Time-Scale and Cost
From: 1995

To: 1996

Cost: £26,529
Contractor / Funded Organisations
Leatherhead Food International Ltd