Chapter 6: Non Parametric Tests Flashcards
Kolmogorov-Smirnov Z
tests whether two groups have been drawn from the same population.
has better power than the Mann-Whitney when samples are less than 25 per group
Kruskal Wallis Test
tests the hypothesis that multiple independent groups come from different populations
compares >2 independent groups of scores
Mann-Whitney
- looks for differences between two independent samples
- tests whether the populations from which two samples are drawn have the same location
- equivalent to the independent t-test
Median Test
tests whether the samples are drawn from a population with the same median
Non-Parametric Tests
used to test hypotheses that don’t make many of the assumptions we have seen
AKA “assumption free tests”
Pairwise Comparisons
compare all pairs of group means
Ranking
how scores overcome the problem of the shape of the distribution
done by giving the lowest score a 1, the nest lowest a 2, etc.
Sign Test
based only on the direction of difference (positive or negative)
magnitude of change is completely ignored
lacks power unless sample sizes are very small (6 or less)
Wilcoxon Rank-Sum Test
looks for differences between 2 independent samples
tests whether the population from which two samples are drawn have the same location
equivalent of the independent t-test
Wilcoxon Signed Rank Test
- looks for a difference between two related samples
- used when there are two sets of scores to compare, but the scores come from the same participants
- equivalent of the related t-test