“…Biased estimates were presented in the result section for two reasons: first, both biased and unbiased estimates revealed identical patterns of results; second, biased estimates are better indicators of the veridical magnitude of the effects of diagnosis and drug treatment on the error types. Barkley et al, 1991Barkley et al, 1989Barkley et al, 1988Coons et al, 1987Fitzpatrick et al, 1992Levy & Hobbes, 1988Matier et al, 1992Michael et al, 1981Milich et al, 1989Pelham et al, 1990Rapport et al, 1987Rapport et al, 1986Taylor et al, 1987Werry & Aman, 1975Yepes et al, 1977 Study Barkley et al, 1991Barkley et al, 1989Barkley et al, 1988Coons et al, 1989Fitzpatrick et al, 1992Levy & Hobbes, 1988Matier et al, 1992Michael et al, 1981Milich et al, 1989Pelham et al, 1990Rapport et al, 1987Rapport et al, 1986Taylor et al, 1987Werry & Aman, 1975Yepes et al, 1977 Size 40 Barkley et al, 1992Barkley et al, 1990Chee et al, 1989Halperin et al, 1992Halperin et al, 1990Hooks et al, 1994Horn et al, 1989Newcorn et al, 1989O'Dougherty et al, 1984Schachar et al, 1988Seidel & Joschko, 1991 Omission…”