Journal of Applied Sciences1812-56541812-5662Asian Network for Scientific Information10.3923/jas.2004.38.42MendesMehmetPalaAkin1200441The present Monte Carlo study compared Type I error rates of four tests under nonnormality and heterogeneity of variance assumptions. Cochran test, Brown–Forsythe test, modified Brown–Forsythe test and approximate ANOVA F test were evaluated for three–and six different groups. At the end of 50,000 simulation trials, the Type I error rates for four tests were affected by the sample size, variance ratio, the number of groups and the relationship between sample sizes and groups variances. Type I error rates of the Brown Forsythe test, extension of the modified Brown–Forsythe test and approximate ANOVA F–test were close to the predetermined alpha levels (0.05), while Cochran tests greatly deviated from the predetermined level.]]>Welch, B.L.,1951James, G.S.,1951Marascuilo, L.A.,1971Brown, M.B. and A.B. Forsythe,1974Wilcox, R.R.,1989Alexander, R.A. and D.M. Govern,1994Levy, K.J.,1978Dijkstra, J.B. and P.S. Werter,1981Tomarken, A.J. and R.C. Serlin,1986Tabatabia, M.A. and W.Y. Tan,1986Lix, L.M., J.C. Keselman and H.J. Keselman,1996Mendes, M.,2002Cochran, W.G.,19542 tests.]]>Asiribo, O. and J. Gurland,1990Mehrota, D.V.,1997Satterhwaite, F.E.,1941Box, G.E.P.,1954Hartung, J. and D. Argaç,2001Anonymous,1994Mendes, M. and A. Pala,2003Krutchkoff, R.G.,1988Oshima, T.C. and J. Algina,1992