Title: Associate Professor
Institution: Washington State University
Street: PO box 643144
Zip-Code: WA, 99164-3144
Authors: Eleanne Solorzano and Nairanjana Dasgupta
Title: Comparing Multiple Treatments to a Positive Control, In the Presence of a Negative Control
Abstract: Routinely experiments in biological sciences include both positive and negative controls. Negative control is often interpreted as ?no treatment?. In other scenarios it is the ?cheapest? or ?easiest to apply? treatment. The Positive control is the ?standard treatment? to which all the competing treatments are to be compared to. In the past the information from the negative control has often not be completely utilized. Often all the competing treatments are compared to the positive treatment ignoring the negative control. In our paper we propose a method that will take into account the information from the negative control, while comparing the treatments to the positive control. The method we propose would compare only the competing treatments that are ?better? than the negative control to the positive control, thus making more efficient use of the Type I error allocation. In finding our critical points we use a combination of ideas from Dunnett (1955) and ! Edwards and Hsu (1983). We provide algorithms for critical point calculation as well as some selected critical points. The idea of this problem came to us from a consulting data set from the Food Sciences Department and we use this as our data example. Our proposed method is compared to Dunnett?s method, which does not account for both controls, via Monte Carlo simulations. Advantages and disadvantages of using the information from negative control are discussed.
References: Dunnett, C.W., (1955). A multiple comparison procedure for comparing several treatments with a control. Journal of American Statistical Association, 50:1096-1121.
Edwards, D. G., and Hsu, J.C., (1983). Multiple Comparison to the Best Treatment. Journal of American Statistical Association, 78: 965-971.