Choosing the correct statistical test by hcj

VIEWS: 52 PAGES: 2

• pg 1
```									                                   Choosing the correct statistical test

1.   What scale of measurement has been used?
2.   Are you testing differences between groups or associations between variables?
3.   Are the samples independent or related (same person tested twice or paired ppl)?
4.   How many groups are involved?

Nominal data:

Nominal Data

Hypothesis of Difference                           Hypothesis of Association

Independent Selection                               Related selection         Any number of cells

Chi Square                                      McNemar Test          Coefficient of Contingency
Ordinal data:

Ordinal Data

Hypothesis of Difference                                                          Hypothesis of Association

Independent Selection                                                                        Related selection                     Spearman r

Two Groups                             More than 2 group                                  Two groups                         More than 2 groups

Mann-Whitney U                        Kruskal-Wallis H Test                              Wilcoxon T test                     Friedman ANOVA by
ranks

Interval/ratio data:

Interval Data

Hypothesis of Difference                                                                  Hypothesis of Association

Independent Selection                                            Related selection                           Two groups                 More than 2 groups

Two Groups                 2 or more groups                Two groups                     2 or more groups              Pearson r                     Multiple R

Independent t Test           One-way ANOVA                    Paired t-test               Within-subjects ANOVA          With prediction               With prediction

More than 1 IV                                                                         Regression of Y on X           Multiple regression

Factorial ANOVA

```
To top