ANOVA 0.0 / 5 ? PsychologyVisual SystemUniversityNone Created by: 10dhallCreated on: 29-11-17 19:31 What does an ANOVA test? The variance in the participants between groups and within groups 1 of 21 What in an ANOVA shows difference/ significance? If the between pp variance is larger than the within pp variance, significance is shown 2 of 21 What can we do to find what is causing the ANOVA difference? Run separate comparisons for each condition comparison or run a t test 3 of 21 What is an ANOVA? Parametric test of variance. Used if there are more than 2 conditions/groups. 4 of 21 What is a one way ANOVA study with 1 factor but more than 2 conditions 5 of 21 How is homogeneity of variance assessed in IG ANOVA? Levennes test - if p>0.05 - equal variance 6 of 21 What is the F score/ratio? BG variance/WG variance - the higher the number, the more difference 7 of 21 What are type 1 and 2 errors? 1= thinking significant result when it wasn't. 2- thinking non significant when it was 8 of 21 What is familywise error? The error rate increases due to interaction between different factors and comparisons - can cause type 1 error 9 of 21 How can we get around this? Post hoc tests - bonferroni correction 10 of 21 Whats the bonferroni correction? divide the acceptable p value by the number of conditions 11 of 21 What is a factorial designs/ multiway ANOVA? With more than 2 IV's 12 of 21 What does factors and levels mean? factors- IV, levels - conditions/groups 13 of 21 What can the reason for difference be in FD? main effects or interactions 14 of 21 How do you interpret interaction graphs? If the lines are parallel then theres no interaction, if they aren't parallel then there is an interaction 15 of 21 What is Maulchlys test of sphericity? RM - more than 2 conditions, it tells us if the conditions can be compared, should not be significant 16 of 21 How would you write the main effect? F(BS df, error df) = F, p=sig, n^2 17 of 21 What is the pairwise comparison output? Post hoc test, comparing conditions to each other 18 of 21 What is a simple effect? Difference between 2 conditions of a factor in a single condition of another factor 19 of 21 How would you test simple effects?? Split the file on SPSS! Use a t-test (RM or IG) - use bonferroni/ other - pos hoc 20 of 21 What is an effect size? strength of relationship or difference between conditions - partial eta squared - z score = score 1-score 2 divided by mean SD 21 of 21
Comments
No comments have yet been made