HSS 2381 Lecture Notes - Lecture 9: Repeated Measures Design, Pooled Variance, Analysis Of Variance
17 Dec 2017
School
Department
Course
Professor
Document Summary
More than 1 data point per person. We are concerned with each participant"s pattern of change, as it varies with different conditions. Also concerned with consistency of changes (lack of interaction between subjects and treatment) Subjects effect not measured, but must be included in source table. F ratio not calculated for s effect. Dirty data: high interaction between subjects and treatment (inconsistent, unpredictable) Clean data: minimal interaction between subject and treatment (consistent results) Error term: interaction of subjects and treatment. When collecting 2 sets of group and you want to groups to each other; 1 measures from the same compare the results score per person group; 2 scores per person with the population. When collecting 2 sets of difference means measures from 2 different between means. Denominator: groups, but each participant estimate of std error in group 1 is matched with. Denominator: of the difference someone in group 2 based on estimate of between group fundamental characteristics population std means.
