What are the two types of t-tests
- Paired (dependent) t-test
2. Independent t-test
Define paired dependent t–test
- Examines differences in one group across two time-points or levels
- No assumption of variance
What does a paired dependent t-test determine
- Test to determine the significance of the difference between two related sets of cores
what does a unpaired t-test determine
Test to determine the significance of the difference between two independent sample means
Define unpaired t-test
Examines differences between two groups
Define degrees of freedom
- Number of individual scores that are free to vary without changing the means
Define homogeneity of variance
- The spread of scores around each mean is approximately equal
Define one-way ANOVA
- A parametric inferential statistical technique for determining if there are differences in the means between three or more different groups/ conditions on a particular dependent variable
How does a one-way ANOVA work
Examines differences for three or more groups
What is used to determine where the difference are
Post hoc
A significant what will tell you there is a difference between groups
F-ratio
Define Repeated Measures Anova
- Analysis of variance in which the independent variable (factor) has been measured on the same participants in all conditions
What do we get from a repeated measures anova
- A main effect for time for repeated measures ANOVA
- If the main effect is significant, follow-up tests (Bonferroni) to see where the change/ differences occurred
What does a repeated measures anova examine
- Examines differences over three or more time-points
Define mixed Model anova (mxed factorial design)
- Several factors some of which are repeated measures (on the same people) and others are between-subjects (om different people)
What do we get from a mixed model anova
- A main effect for group
- A main effect for time
- Potentially an interaction effect
What does a mixed model anova examine
- Examines differences between groups and over time
What is Benferroni correction and when to use it
- Type 1 error, increase chance of falsely saying there was a difference
- Need to use Bonferroni correction
- 0.05/4 = 0.0125
- So, to be a significant difference we need p<0.0125
