Abstract
In PLT, it is often useful to divide the test [human] cells for any given reagent into a positive and a negative group. Some general problems in PLT analysis were discussed. A model for PLT data is described to provide a conceptual framework. A statistical procedure based on the model was suggested.