ANOVA Flashcards
Simple Hypothesis
µ1=µ2=µ3
Complex null hypotheses
(µ2+µ3)/2=µ1, (µ1+µ3)/2=µ2, (µ1+µ2)/2=µ3
Omnibus test
Evaluates all of the simple and complex null hypotheses at the same time
Post-hoc tests
If we reject the null hypothesis with the omnibus test, this determines which of the null hypotheses are responsible.
Tukey Test
Most widely used Post-Hoc test. It identifies which of the pairwise mean comparisons might be responsible for the rejection of the null hypotheses in the Omnibus test.
What if the tukey test fails?
If Omnibus=reject null,& if tukey fails, then one or more of complex hypotheses is responsible. Tukey only compares μ1 = μ2 = μ3, not complex.
What does ‘=’ notation mean?
the means are ‘about the same’ or that we do not have enough evidence to reject them as being the same.
What happens to the graph when two groups are averaged together in a complex hypothesis?
The width of the average distribution goes down and it is easier to reject a value as coming from the average distribution than from either individual distribution.
Apriori Hypothesis
Tests of a specific complex null hypotheses
Apriori v. Omnibus
Apriori has an advantage over the Omnibus test in that they are easier to reject due to less variance. Similar to a one-tailed test.
Anova
Uses variances estimates to test the null hypotheses of the Omnibus test
How to get the Estimate of Population Score Variance
- by looking at the scores within each group. If we have 3 sample estimates, we can combine these estimates to get out best estimate of the population variance.
- means of each of the groups: mean distribution variance is N times smaller than score distribution variance.
MSWithin
The variance estimate coming from scores within the groups.
MSBetween
The variance estimate coming from means between the groups
F Ratio
Compares the two variance estimates through a ratio by dividing MS Between/ MS Within
When do you use a simple ANOVA and why is it valuable?
It is used when the number of scores in each group is equal. It is valuable because it is both easier to understand and easier to calculate than the structural ANOVA
When do you use a structural ANOVA?
It can be used in all cases, when the scores in each group are equal or if they are not equal.
If the null hypothesis is true, what happens to the variance of means?
It should be relatively small. It should be about the same as the variance of the scores. F ratio is about 1 because MSB and MSW are about the same.
If the null hypothesis is false, what happens to the variance of the means?
The variance of the means will rise dramatically because now the means are much further away from each other because they are coming from different distributions. F ratio is greater than 1 because MSB increases while MSW stays about the same.
If the null hypothesis is false, what happens to the F ratio?
The ratio of MSB / MSW will also rise because the numerator (MSB) will increase while the denominator (MSW) will stay the same.
When do we use the One-way between subjects ANOVA?
The ratio variables are not related and there are 3 or more groups.
1-way between subjects simple ANOVA - Score variance
Sj2=Σ(Xi-X)2/(n-1)
1-way between subjects simple ANOVA - MSwithin
MSWithin=ΣSj2 / K
1-way between subjects simple ANOVA - MSbetween
MSBetween= n * Σ(Xj - XG)2/(k-1)
1-way between subjects simple ANOVA-F obs
MS Between/ MS Within
1-way between subjects simple ANOVA-DFTotal
dftotal = Ng-1
1-way between subjects simple ANOVA- DFBetween
df between= k-1
- remember that n is the number of scores per group
- k is the number of groups
1-way between subjects simple ANOVA- DFWithin
df Within=(n-1)k
- remember that n is the number of scores per group
- k is the number of groups
1-way between subjects structural ANOVA
Must be used when the sample sizes are unequal, but also can be used if the sample sizes are equal. Comparing an estimate from the mean variance and an estimate of score variance. for EVERY score in all groups, there is a total amount of variability associated with that score.
1-way between subjects structural ANOVA- Sum of Squares Within
Σ(X - Xj)2
1-way between subjects structural ANOVA- Sum of Squares Between
Σ(Xj - XG)2
1-way between subjects structural ANOVA- Sum of Squares Total
Σ(X - XG)2
1-way between subjects structural ANOVA- DF Between
K - 1
1-way between subjects structural ANOVA- DF Within
NG-K