RANK DETERMINATION OF SPECTROSCOPIC PROFILES BY MEANS OF CROSS-VALIDATION - THE EFFECT OF REPLICATE MEASUREMENTS ON THE EFFECTIVE DEGREES OF FREEDOM

In the present work the effect of sample replication and 'true' instru mental resolution on the number of significant latent variables obtain ed by cross validation is investigated. Three sample sets containing t hree and four chemical components were prepared according to mixture d esign. The samples were analysed by Fourier transform infrared spectro scopy, and cross validation was performed using the predicted residual error sum of squares (PRESS) algorithm with a varying number of degre es of freedom. It was found that the correct number of chemical compon ents could be estimated precisely from the instrumental data (the X ma trix) presuming that the degrees of freedom were corrected for the abo ve-mentioned effects.