fahdguthmy
New member
- Joined
- Jan 3, 2020
- Messages
- 14
Hello.
The Kruskal Wallis and the Median Tests are used when the distribution of the data does NOT follow a normal distribution. But the book I'm reading tells me otherwise.. Here is the extract that sums up my confusion:
"Kruskal-Wallis test. This is one of the most popular nonparametric tests for comparing K independent samples. It is the nonparametric analog of one-way ANOVA. In p value calculations, mid-ranks are substituted for the raw data and exact permutational distributions are substituted for F distributions derived from normality assumptions. It has good power against location-shift alternatives, where the distributions from which the samples were drawn have the same general shape but their means are shifted with respect to each other. It is about 98% as efficient as one-way ANOVA for comparing K samples when the underlying populations are normal and have a common variance."
What's an F distribution? And how come its normal when the data isn't?
The Kruskal Wallis and the Median Tests are used when the distribution of the data does NOT follow a normal distribution. But the book I'm reading tells me otherwise.. Here is the extract that sums up my confusion:
"Kruskal-Wallis test. This is one of the most popular nonparametric tests for comparing K independent samples. It is the nonparametric analog of one-way ANOVA. In p value calculations, mid-ranks are substituted for the raw data and exact permutational distributions are substituted for F distributions derived from normality assumptions. It has good power against location-shift alternatives, where the distributions from which the samples were drawn have the same general shape but their means are shifted with respect to each other. It is about 98% as efficient as one-way ANOVA for comparing K samples when the underlying populations are normal and have a common variance."
What's an F distribution? And how come its normal when the data isn't?