-
t-test
- For continuous data
- Normally distributed
-
2 types
- 1.For independent measures
- 2.For Repeat measures
- Only for TWO GROUPS
- As multiple tests could lead to the wrong conclusion
-
ANOVA
- For normally distributed data
- More than Two groups
- Can identify an interaction
-
Sign test
- For dichotomous data
- Non Parametric
-
Friedmans
- Non-parametric data
- Repeat measures
-
Kruskall wallis
- Non parametric data
- Independent measures
-
Mann-Whitney U test
- Non parametric continuous data
- Ordinal data
- Independent design
-
Wilcoxon
- For repeat measures
- Ordinal data
- In place of paired t-test when data is non-parametric
-
Chi-squared
- For nominal data
-
Way to assess a piece of research
-
Identify the following
- IV
- DV
-
Research design
- Independent/repeated?
- Type of data obtained
- What test to use to analyse it
- What stat is generated
- How to interpret the stat generated