GPHY 247 Lecture Notes - Lecture 12: Type I And Type Ii Errors, Complement Factor B, Analysis Of Variance
10 Apr 2016
School
Department
Course
Professor
Document Summary
11. 1 two-factor anova: explanation of two-factor anova. 3 kinds of questions: is there an effect of factor a. A. between 2 rows: is there an effect of factor b. Between 2 cells within a column or a row: two-factor anova visualization. Graphing along a single fact makes it dif cult to see the main effects and interaction. Better to graph an interaction plot. x-axis is one factor, symbols or lines are the other. 11. 3 two-factor anova assumptions: evaluating assumptions qualitatively. Residuals are normally distributed (between each cell) Residual variance is the same for all cells (shown by the width of the distribution) Y-values represent random and independent observations: evaluating assumptions quantitatively. Residuals are normally distributed: shapiro-wilks test ho: residuals are normally distributed. Residual variance is the same for all groups: bartlett"s test ho: all groups have equal residual variance. Ha: at least 2 groups do not have equal residual variance. 11. 4 two-factor anova contrast statements: testing all contrasts.
