Application of Statistical Selection Procedures in Biotechnology

* Final gross prices may vary according to local VAT.

Get Access


A short methodical summary about the two classes of statistical selection procedures is given, i.e. the indifference zone (and d-correct) procedures and the subset procedures. The biotechnological problem consisted in selecting the „best“ or at least a „good“ mutant with high enzyme activity from a set of eight mutants of the species Aspergillus niger with a large probability. Depending on suppositions about the variances of the enzyme activities, different selection rules are applied. Starting with the subset procedure of Gupta (1985) for the case of equal variances, the number of mutants is reduced. The following d-correct procedure of Bechhofer, Dunnett and Sobel (1954) calculates the necessary sample size n. Then the mutant whose sample has the largest mean will be selected as a „good“ one with a given precision of d and a probability of correct selection of (1-β). As long as the variances of the enzyme activities are assumed to be different, the selection procedure of Dudewicz and Dalai (1975) must be used.