CuSO4 Hb screening gives large numbers of false passes, particularly in women. Using venous samples, the majority correctly pass at the lower HemoCue cut-offs. The current dual-testing policy appears convenient for donor sessions, but because small percentages of false passes and fails represent large numbers of donors, every effort should be made to improve the accuracy of Hb screening.

No abstract

Extends the theory presented in Smith et al. (1975) to larger confusion matrices. Stimuli and responses are described by a series of features which take a discrete set of values. Processing models are proposed where each feature is tested independently (Independence model), where the processing of a given feature depends on the responses that have been made in testing for other features (Response Conditional model) or where the processing of a given feature depends on whether other features have been correctly identified or not (Correct Conditional model). Computer programs are described which fit these models to arbitrarily large confusion matrices. Difficulties in interpreting the behaviour of these models with real data are reviewed and solutions are offered. These points are illustrated by detailed analyses of confusion matrices obtained in studies of speech perception, memory for speech and semantic memory. order of feature-testing units is considered, and x2 computed for each order. The program prints out the x2 associated with each testing order, and the testing order that leads to the smallest x2, together with the parameters associated with the minimum x2 testing order and the predictions derived from these parameters. CONFUSION MATRIX ANALYSIS. I1 59 1 0 16 s2 9 48 7 12 s3 5 6 53 11 s4 11 12 11 5 2 32 SMITH AND JONES [Br. r. nrath. statist. Psychol.