User Tools

Site Tools


help:kruskal

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

help:kruskal [2010/12/20 02:37] (current)
Line 1: Line 1:
 +[[http://​www.sofastatistics.com/​userguide.php | Contents]]
  
 +[[:​help:​stats_tests | Statistical Tests Available]]
 +
 +====== Kruskal-Wallis H ======
 +
 +[[http://​en.wikipedia.org/​wiki/​Kruskal%E2%80%93Wallis_one-way_analysis_of_variance | Kruskal-Wallis H]] lets you decide if there is a difference in an ordinal variable e.g. user satisfaction ratings, between groups. ​ It is the non-parametric equivalent of the [[:​help:​anova | ANOVA (Analysis of Variance)]] relying on value ranks rather than data values themselves.
 +
 +Because it is based on order rather than value, the output table displayed by SOFA shows the median instead of the mean.
 +
 +The p value tells you how likely the difference observed would occur if drawing from the same population (i.e. where the groups didn't really have a relationship with the value being averaged). A small p value tells you that it would be rare to observe such a difference if the group doesn'​t have a relationship with the value being averaged. ​ From this we might reject the null hypothesis in favour of the alternative hypothesis - namely, that there is a difference according to the grouping variable.
 +
 +In the example below (based on false data for illustration only), we shouldn'​t be surprised the p value is very low.  It is possible to see a large difference in average age and N is reasonably large for each group. ​ In this case we could reject the hypothesis that nation has no relationship with age.
 +
 +{{:​help:​kruskal_wallis_h_example.gif|}}
 +
 +Be aware that there is a certain sensitivity about terminology around this area.  According to a widespread convention, we shouldn'​t conclude that there is a relationship,​ only that we reject the null hypothesis (see [[http://​www.stats.gla.ac.uk/​steps/​glossary/​hypothesis_testing.html#​h0 | Hypothesis testing]]). We might go so far as to reject the null hypothesis in favour of the alternative hypothesis. ​ See [[http://​en.wikipedia.org/​wiki/​Statistical_hypothesis_testing | Statistical hypothesis testing]].
 +
 +If your data is not numeric and it is adequately normal the [[:​help:​anova | ANOVA (Analysis of Variance)]] is a good alternative.
 +
 +[[http://​www.sofastatistics.com/​userguide.php | Contents]]
 +
 +[[:​help:​stats_tests | Statistical Tests Available]]
 +
 +[[:home | Wiki]]
help/kruskal.txt ยท Last modified: 2010/12/20 02:37 (external edit)