Lecture 14 - Non-Parametric Tests Flashcards
When can parametric tests not be used?
When data is not normally distributed.
When ordinal data is used.
When data are heavily skewed - the step size between points on the scale is probably not constant.
What is the non-parametric equivalent of Pearson’s r?
Spearman’s R
What is the non-parametric equivalent of a paired-samples t-test?
Wilcoxon T
What is the non-parametric equivalent of an idependent-samples t-test?
Mann-Whitney U
What is the non-parametric equivalent of a repeated-measures ANOVA?
Friedman’s
What is the non-parametric equivalent of a between-groups ANOVA?
Kruskal-Wallis
In what situation is a Mann-Whitney U test appropriate?
When the data are non-parametric, the design is between-groups, there is one IV with 2 levels, and the DV is on at least an ordinal scale.
What is the rationale behind a Mann-Whitney U test?
When all the data is arranged in ascending order, the groups are either clustered (H1) or randomly distributed (H0).
How is a Mann-Whitney U test conducted?
All scores are ranked and the rank totals for each IV level are calculated and compared.
The test statistic (U) is calculated:
U = n1n2 + (nx(nx+1))/2 - Tx
Where n1/2/x= no. of pts in grp 1/2/larger rank total and Tx= total larger rank
if Uobserved<=Ucritical, H0=rejected.
How do you conduct a Mann-Whitney U test in SPSS?
Analyse/Nonparametric Tests/2 independent samples
Test variable = DV
Grouping variable = IV (define groups)
Check Mann-Whitney U box.
In a Mann-Whitney U test SPSS output, what output tables are there and which values are useful?
Ranks and test statistics. Sum of ranks is useful, but the values you need are the Mann-Whitney U and the Asymp. sig., which is the p value.
A Wilcoxon T test is also known as…?
Wilcoxon related-samples test
Wilcoxon signed-rank test
The Wilcoxon signed-rank test is not to be confused with the…?
Wilcoxon Rank Sum test, which is more like Mann-Whitney U.
What is the rationale behind a Wilcoxon test?
It is based on the difference between the two scores for each subject. It uses both the direction and magnitude of the differences (using ranks, so it can be used on skewed data). If H0 is true the differences in each direction will be the same magnitude.
How is a Wilcoxon test performed?
The differences and their direction are calculated and ranked regardless of direction. The totals for negative and positive (T- and T+) are calculated.
The test statistic (W) is the smaller of the T values. H0 can be rejected when Wobserved<Wcritical.
