Bivariate inductive statistics - interval and nominal variables - dependent groups

Two-Way analysis of variance (One-Way repeated measures ANOVA)

ANOVA compares the means of three or more groups (group represents level of factor) to determine if at least one group mean is different from the others. The F-ratio is used to determine statistical significance. The tests are nondirectional in that the null hypothesis specifies that all means are equal and the alternative hypothesis simply states that at least one mean is different. Randomized blocks must be analyzed using 2-Way ANOVA. The randomized block design takes account of known factors that affect outcome but are not of primary interest.

The two steps in randomized block design are:
1. Collect together homogeneous experimental units (e.g. people of same age, sex) into a block.
2. Assign treatments at random to the experimental units within a block.
Block often represents one person, who takes different medicine (treatment). In the analysis treatment is fixed factor and block is random factor. Data for the analysis must be entered so that one column is response variable, second is fixed factor and another random factor. One row represents one observation. Design with 3 treatments and 5 blocks must have exactly 3 x 5 = 15 rows.