1)One way ANOVA is the appropriate statistical test when there is one nominal independent variable with at least 3 levels; one sacle DV, and either between groups or within groups design 2)One way ANOVA
Null Hypothesis: No differences between population means.
Alternative Hypothesis: At least one pop mean is different from at least one other pop mean. (Can’t use symbols) 3)Numerator of the F statistic measures between groups variance (MSbetween) 4)Denominator of the F statistic measures within groups variance (MSwithin) 6)A priori test: planned ahead of time, before you collect data decide on test, based on reasoning Post hoc: choose after you look at data; based on data, choose groups you want to compare; usually harder to find significant difference with a post hoc test than an a priori test 7)Tukey HSD tells you which group is significantly different from the others. 8)Effect size (R2) tells you proportion of variance in the DV that is accounted for by the IV Small: 0.01
9)Repeated measures ANOVA is appropriate when you have a within groups design with one IV. 10)
In repeated measures ANOVA, can’t calculate SSwithin directly, need to calculate SSsubjects and then subtract SS¬between and SSsubjects from SStotal to find SS¬within. Makes SSwithin smaller. 11)
MSwithin will be smaller because take out variability due to subjects. Means F-value will be bigger and makes it more likely to reject null. Much more powerful test then between groups ANOVA.
Assumptions: One way between groups ANOVA
•Random selection of samples
•Normal distribution of DV in pop
•Homogeneity of variances
(samples come from pop with similar variances)
One way repeated groups ANOVA
•Same assumptions as above
•Control for order effects
Two way ANOVA
•Same as above
Two way ANOVA is appropriate statistical test when you have more than one IV, each with at least 2...