Howard Mark (hlmark)
Senior Member Username: hlmark
Post Number: 231 Registered: 9-2001
| Posted on Monday, June 01, 2009 - 9:48 am: | |
Estefania - the number of samples should be the same as you would use for any validation study. But if you are going to the trouble of setting up a true blind validation test, then this project is probably more critical than the average. Under these circumstances, while the NUMBER of samples is the same, the choice of WHICH samples to include in the validation set should be more carefully chosen than you might otherwise use, to make sure that the validation sample set adequately covers the multivariate space (in chemometrics-speak), or in other words, includes samples with all combinations of the sample's variations (high-low, high-high, etc., values of the consituents, in English-speak). \o/ /_\ |