One way within subjects anova Flashcards
when do you use a one way within subject anova?
single within subjects variable with 3+ levels
what test would you use if it was two levels
related samples t-test
what is the non-parametric equivalent
Friedman
what is the manipulation involved in one way within subjects another
within subjects difference between subject differences e.g. ind diffs chances factors I.e. error variance
what does a one way within test?
the diff between within subjects variance is greater than error variance after controlling for between subjects variance
assumptions of one way within subjects
parametric - normally distributed
ratio / interval / ordinal data
sphericity
(even if some violated, ANOVA still used as best option in terms of power)
what is sphericity
equality of variances in differences between levels of the IV
e.g. A-B Var 15.75
A-C 16.42
B-C 16.01 = all relatively similar
if we want sphericity then what do we want the test to show
non significant
what do you use to illustrate the data in these anovas
simple line graph
error bars
what test do we use if we violate the sphericity assumptions
Huynh - fendt statistic