Non-Parametric Tests Two Flashcards
what would you use to analyse correlations
P - pearsons r
NP - spearmans r
what would you use to analyse 2 samples
P - paired samples or unpaired samples
NP - wilcoxon t or mann whitney u
what would you use to analyse k samples
P - rep measures anova or between groups anova
NP - friedmans or kruscal-wallis
what is a k sample
a systematic sample of a group
describe the kruskal wallis h test
non parametric
looks at differences between conditions in between groups designs
used when theres multiple levels - 1 IV with more than 2 levels
requires DV to be ordinal
how do you do a kruskal wallis H test in spss
analyze
nonparametric
k independent samples
clicl kruskal wallis h
how do we interpret kruskall wallis tests and friedman tests
SPSS converts h to chi squared
x2 = ………, df=……., p=0.05
it only tells us that the three conditions differ, but not which ones!
to work out which ones we use multiple mann whitney tests (for k) or wilxocon tests (for f)and apply the bonferroni correction
how to we interpret an anova test
f=(DF1, DF2) = F
p 0.05 (p=SIG)
how do we interpret a t test
t=(DF)=T
p 0.05 (p=SIG)
how do you do a bonferonni correction
multiply the p value by the possible number of tests
describe the friedman test
nonparametric
designed for paired samples, both within and mathched subject designs
used when theres 1 IV with multiple levels
requires DV to be ordinal
how do we do a friedman test in SPSS
analyse
nonparametric
krelatedamples
click friedman