Factorial ANOVAs Flashcards
based on analysis that is based on 2 IVs (factors) or more =
2 way factorial ANOVA
how do factorial ANOVAs allow us to look at the interaction between factors?
they test the differences between conditions for more than 1 categorical independent variable
mean difference is constant across all levels of factor (IV) combinations =
no interaction
different combinations of factors and levels give different means =
interaction
the pattern of responses for each level is different depending on the factor it is paired with. this is called an _______
interaction
factorial ANOVAs produce a ____ _____ for each factor as well as an interaction
main effect
outcome that can show consistent difference between levels of a factor =
main effect
what do main effects indicate?
whether there are significant differences between the levels of 1 factor
what does the output for a 2 way factorial ANOVA include?
2x main effects, 1x interaction, mean values for each group (these means describe the interaction)
what is generated from each main effect?
a marginal mean
what do marginal means tell us?
which level of each factor resulted in better performance
where do you find the marginal mean in the SPSS output?
in the ‘Mean’ box
what table has the values you need to report the main effects and the interaction?
Tests of Between-Subjects effects
for each factor how do you report the main effect and the interaction between the 2 factors?
F(effect df, error df) = [F value], p = [p value], np2 = [np2 value]
if there is a significant main effect you need to include the _____ in the write up
means
if there is a significant interaction what is required?
follow up tests called SIMPLE EFFECTS TESTS
what are simple effects tests?
basically bonferroni corrected t tests
what stats are reported from simple effects tests?
mean, SD, t value, df, p value
how do you calculate the bonferroni correction?
divide alpha 0.05 by the number of t tests you do
what else is needed in the report of simple effects tests?
a table of means
list the steps to run the descriptive stats for a between factorial design
DATA > SPLIT FILE > ORGANISE OUTPUT BY GROUPS. put one of the factors into the box ‘groups based on’. ANALYSE > DESCRIPTIVE STATS > EXPLORE. move DV to dependent list and move factor that was not moved to ‘groups based on’ into the factor list. before running the data you must unsplit it again: DATA > SPLIT FILE > RESET > OK
list the steps to run a factorial ANOVA
ANALYSE > GENERAL LINEAR MODEL > UNIVARIATE. move DV to ‘dependent variable box’ and both factors to ‘fixed factors box’. select plots. move 1 factor to horizontal axis and other into separate lines box > add. repeat this but swap the factors around > continue. choose correct ‘options’. select EM means. move 2 factors to ‘display means for’ box > continue > ok.
what options need to be selected for the factorial ANOVA?
descriptive stats, estimates of effect size, homogeneity tests
what does the ANOVA produce in the SPPS output?
descriptive stats, Levene’s test of error variances, table of between subjects effects, estimated marginal means for each factor (main effect) and 2 line plots
list the steps in running simple effects tests
DATA > SPLIT FILE. organise output by groups. move 1 factor across to ‘groups based on’ > ok. ANALYSE > COMPARE MEANS > INDEPENDENT SAMPLES T TEST. move other factor that wasn’t split into grouping variable. define groups > 1+2. DV into test variables. then reset split data file and swap factors around. (repeat this whole process with the factors also swapped around)
