4. Choosing a Test

Question 1

Taking measurements and looking for differences between groups

Answer 1

One sample r test:
Sign test

Two samples:

1. ) paired r-test if matched pair sample -> Wilcoxon matched pairs
2. ) Two sample r test if no matched pair sample -> Mann-Whitney U test

More than two samples:

1. ) effect of one factor (One Way Anova) -> Kruskal-Wallis test
2. ) effect of two factors (Two-way of Anova)

Question 2

Sign test

Answer 2

statistical test to compare the sizes of two groups. It is known as a non-parametric or distribution free test, meaning the data isn’t from a normal distribution. It tests the null hypothesis that the median of a distribution is equal to some value.

It is an alternative test to the one sample t-test or a paired t-test.

Question 3

1. ) Wilcoxon matched pairs test

2. ) Mann-Whitney U test

Answer 3

1.) non-parametric test comparing two groups.

2.) Comparing differences between two independent groups when the dependent variable is either ordinal or continuous, but not normally distributed.
Often considered the non-parametric alternative to the independent t-test (not always the case)

Question 4

1. ) One-way Anova -> Kruskal Wallis test

2. ) Two-way Anova

Answer 4

1.) One-way Anova:
type of statistical test that compares the variance in the group means within a sample. Only one independent variable or factor is being considered. Hypothesis based test, meaning it aims to evaluate multiple mutually exclusive theories about the data.
Kruskal Wallis test:
Tests whether samples originate from the same distribution. It can compare two or more independent sample sizes.

2.) tests the effect of two nominal predictor variables on a continuous outcome variable.
Anova -> Analysis of variables on independent variable.

Question 5

Counting frequencies in different categories

1. ) expected outcome
2. ) Association

Answer 5

1. ) X2 test for differences

2. ) X2 test for associations