# rmc

HideShow resource information
• Created by: lewis
• Created on: 08-12-15 18:26
Experimental design..
systematically varied IV on DV.
1 of 22
Quasi-experimental design..
uses variables we cannot directly manipulate. Harder to infer causation/C&E. (natural)
2 of 22
Correlation design..
No DV and IV.
3 of 22
Simple comparisons..
One IV only, 2 conditions (either between or within)
4 of 22
One way designs..
One IV only, 3/+ conditions in one experiment.
5 of 22
Factorial designs..
More than one IV, equivalent to 2 experiments simultaneously.
6 of 22
Rationale..
presented in introduction, reason for study.
7 of 22
ANOVA is..
an analysis of variance.
8 of 22
ANOVA compares..
within group error variance to between group variance.
9 of 22
If variance within each group is similar to variance between groups..
the behaviour measures is the same in all groups.
10 of 22
F-ratio..
between groups variance / within groups variance.
11 of 22
What does an f-ratio test?
the likelihood of two variance estimates being from the same population.
12 of 22
What does a significant ANOVA tell us?
somewhere in analysis at least one group is different from at least one other group.
13 of 22
Planned comparisons (a priori)
decide what differences we want to look for.
14 of 22
Post hoc ( a posteriori)
check if there are any differences.
15 of 22
Problems with multiple comparisons..
Family-wise error rate.
16 of 22
Type I error..
conclude there is a difference when there isnt. 5% chance.
17 of 22
Type II error..
conclude there is not a difference when there is.
18 of 22
Non parametric tests..
Few/no assumptions about distribution. Ordinal (ranked) data. Greater chance of type II error.
19 of 22
The Kruskal-Wallis test..
similar to Mann U. Ordinal (ranked) data. Compares 3+ groups.
20 of 22
Kruskal Wallis test is what type of ANOVA?
Non parametric independent one-way ANOVA. Tests for differences betweeen 3+ treatment conditions using a separate group for each treatment.
21 of 22
Friedmans test is what type of ANOVA?
non parametric repeated measures one-way ANOVA. Tests for differences between 3+ treatment conditions using the same group for each treatment.
22 of 22

## Other cards in this set

### Card 2

#### Front

Quasi-experimental design..

#### Back

uses variables we cannot directly manipulate. Harder to infer causation/C&E. (natural)

### Card 3

#### Front

Correlation design..

### Card 4

#### Front

Simple comparisons..

### Card 5

#### Front

One way designs..