A monte carlo comparison of several methods for analyzing small sample binomial data in a disproportionate two#x2013; factor experiment without replication

摘要

Four methods of analyzing binomial data were compared for the case of a disproportionate 3x3 design without replication. With respect to empirical signigicance levels and power, the two-way ANOVA if raw 0,1 cell data is recommended as the preferred method of analysis.