# Statistical Sciences 2244A/B Lecture Notes - Type I And Type Ii Errors, Analysis Of Variance, Observational Error

82 views8 pages

30 Jul 2013

School

Department

Professor

## Document Summary

Simplest experimental design: 2 groups (1 experimental and 1 control, = 1 iv, 2 levels, and 1 dv. Simplest statistic for 2 groups = t-test if increase number of groups (3 or more), then build onto the design. Move from the t-test statistic to the analysis of variance (anova: the anova design is defined using the prefix way (1 iv 1. Way , 2 iv 2 ways ) for example, if: 1 iv, 1 dv and 3 or more groups (eg. 3 levels of the iv), then design is a 1-way anova, iv a . A3: example is a 1x3 anova design, results in 3 cells, cells = iv condition combinations. In a one-way anova, question is whether or not one of the cells is different from the others, termed a main effect . Ho: m1 = m2 = = mk. Ha: at least 1 mean is different. (controls type i error rates as not doing multiple t-tests)