Wilcoxon signed Rank
- paired data
- non parametric data
Mann -whitney
- unpaired data
- non parametric data
unpaired t test
- parametric
- unpaired t test
paired t test
- paired data
- parametric data
What do we do when we want to compare results of studies involving more than 2 groups?
- one- way ANOVA test which tests the null hypothesis that all groups contain random samples from the same population (i.e no statistically significant difference exists between any of the groups). The p value output can then be compared to the chosen alpha.
Why can’t we just conduct multiple t-tests?
- increases probability of false positives occurring (type 1).
What is the problem with ANOVA test?
- only indicates if there IS a statistical difference between the two groups but it does not indicate which groups are different.
=> eg: 20 drugs testing compares each drug to the control and not with each other.
What is the solution to on way ANOVA- to test statistical significance of specific pairs of groups ?
-pairwise comparison, ‘post-test’ conducted in the event of a significant ANOVA result.
What are different types of ‘post-test’
- Tukey’s test : used when a study requires pairwise comparison of every possible combination of groups
- dunnett’s test: used when each pairwise comparison involves one specific group
- Bonferroni’s test : used when the specific pairwise comparisons required do not follow a particular pattern
one way ANOVA used for
- unpaired data
- parametric data
Repeated measures ANOVA used for
- paired
- parametric
Kruskal -Wallis test used for
- unpaired data
- non parametric
Friedman test is used for
- paired
- non parametric data
When investigating the effect of 3 or more varying conditions on a single continuous variable you use….
one way ANOVA
When investigating whether there is an interaction between two categorical variables on a single continuous variable you use…
two way ANOVA
