Multifactor Analysis of Variance.pdfVIP

Multifactor Analysis of Variance Two-way ANOVA The two-way ANOVA is an extension of the one-way ANOVA, except there are two independent factors (hence the name two-way). If groups are formed along more than one dimension, differences among means are attributable to more than one source. Consider an example with six groups, three of women and three of men in which the DV is scores on a final examination in statistics. One source of variation in means is due to gender (SSG). If the three groups within each gender are exposed to three different methods of teaching statistics, a second source of differences among means is teaching method (SS ). The final source of known T differences among means is the interaction between gender and teaching methods (SSGT), which assess whether effectiveness of teaching methods varies with gender. I. Factorial designs a. Factorial designs allow us to measure two different sorts of factor effects. i. The main effect of each factor is the effect of each factor independent of the other factors. ii. The interaction between factors is a measure of how the effects of one factor depend on the level of one or more additional factors. II. Null hypotheses a. There is no main effect for factor A. b. There is no main effect for factor B. c. There is no interaction between the two factors. III. Models for two-way ANOVA a. Model I b. Model II c. Mixed model IV. Two-way ANOVA table Two-way ANOVA with replication Sum of Degrees of Mean F Test squares freedom Squ