Nonparametric Models Flashcards
When do you use nonparametric models?
when the assumptions of parametric tests haven’t been met
statistic
sample
parameter
population
what does ranking help with
outliers, skew
what are the non-parametric equivalents of the independent t-test
wilcoxon-rank sum
mann0whitney
what assumptions are we concerned about with small n
normality, homogenity of variences
small effect size
0.1
medium effect size
0.3
large effect size
0.5
how do you rank data
find the lowest score, give it a score of 1, find the next lowest, give it a score of 2, etc
purpose of wilcoxon rank-sum and mann whitney tests
use when you want to compare distributions in 2 conditions and these conditions contain different entities and the resulting data has unusual data or violates assumptions
fxn of kolmogorov-smirov z
tests whether 2 groups have been drawn from the same population
fxn of moses extreme reactions
compares variability of scores in 2 groups
what test is the moses extreme reactions similiar to?
levene’s test
fx of the wald-wolfowitz runs
looks for runs of scores from the same group w/in the ranked order
what do you need to report for mann-whitney test results
test statistic (U), effect seize, exact values of p
test statistic T
from Wilxocon signed-rank test
sum of positive ranks
what do you report in wilcoxon signed-rank test?
test statistic (T), exact significance, effect size
what does it mean if wilcoxon signed-rank test statistic is <0.05
the 2 conditions are significantly different
when you rank data in wilcoxon signed-rank what do you do with the sign?
ignore it
when do you use the kruskal -wallis test
comparing several independent groups when assumptions have been broken
what can the kruskal-wallis be considered a nonparametric equivilant of?
one-way independent ANOVA
What do you report for kruskal-wallis test?
test statistic H, degrees of freedom, significance, effect size
what does jonckheere-terpstra test do
tests for an ordered pattern to the medians of the groups you’re comparing
step-down methods
identify groups that are the “same” (homogenious)
