BETWEEN SUBJECTS/ INDEPENDENT ONE WAY ANOVA Flashcards
when do we use a between subjects/independent one way ANOV?
groups must be independent- one ppts is in one condition
Normally 3 or more levels in the IV condition (for example time of day= morning, evening, afternoon and no ppts is in more than one of these conditions)
How do we test for homogeneity
look at the ‘levenes test of equality of error variances’
Find the sig value on the based on Mean row- that is your P score
We want P to be NON SIGNIFICANT
if its more than 0.05 then it is non significant and there is no homogeneity
How do we formally write up an ANOVA?
report the homogeneity
then do a table for the means and SD of each IV
Then report if its significant F(df (IV), df(error))= F(IV), p<sig (IV)
then report the effect size (partial eta squared on the IV row) and report if its size according to Cohens H2 criteria
what is PLANNED/APRIORI comparisons
Researcher has a previous hypothesis for a comparison between 2 groups (hence its called planned)
Conducted when the researcher has hypothesised which means will differ from
each other in advance.
–The overall main effect does not need to be significant to run planned
comparisons.
How to report Planned/apriori comparisons
look at the contrast test table on the assumed equal variances rows- find the T, DF, sig (2tailed) scores.
Look at the contrast effect sizes and report the point estimate for cohens d (this is your effect size)- figure out its size using cohens D guidelines.
Reporting it = t(df score)= T score, p<sig score (2tailed)
for example t(42)= 3.98, p<0.001
Then go back to descriptive stats to see which is larger and state the relationship
Report the effect size, and its size
EXAMPLE- The mean recall scores of the morning group were significantly higher than the
evening group, t (27) = 4.135, p<.001, with a large effect (d= 1.80).
note- we know the morning group was ‘higher’ by looking back to the descriptive statistics
