RDA revision

HideShow resource information
  • Created by: Ellis
  • Created on: 18-05-16 10:14
Variance
Mean of all squared deviations from the mean.
1 of 24
Standard Deviation
Square root of variance and is a measure of dispersion
2 of 24
Normal distribution
Majority of measurements are 68%, which is +/- 1 SD
3 of 24
Z-scores
Produced using deviation from the mean= score/SD
4 of 24
Type 1 error
Thinking there is a genuine effect in our population when there isn't. A probability alpha (0.5)
5 of 24
Type 2 error
Occurs when we believe there is no effect on the population, but there is one. A Beta level (0.2)
6 of 24
Family-wise error
Probability of making one or more false discoveries (type 1) among all hypothesis
7 of 24
ANOVA
Tests is 3 or more sample means come from the same population
8 of 24
ANOVA assumptions
Homogenity of variance. Data in interval scale. Sampling distribution of means is normal
9 of 24
Types of ANOVA
Between groups ANOVA. Repeated measures ANOVA. One-way ANOVA
10 of 24
Mauchly's test of sphercitiy
One relevant to repeated measures ANOVA. Asks if data meets assumption that the correlations between all groups is roughly equal. Sig if smaller than 0.5, but assumption has been violated
11 of 24
Post-hoc tests
Consist of pairwise comparisons, are explore the data for any between-group differences. Bonferroni.
12 of 24
Bonferroni correction
Ensures type 1 error is below 0.5
13 of 24
Partietal eta squared (np2)
For between groups ANOVA and repeated measures ANOVA. Creates an estimate of effect size
14 of 24
Non-parametric alternatives to ANOVA
Friedman's test- within subjects design- If sig. use Wilcoxon with a bofferoni correction. Krushall-Wallis test- between subjects design
15 of 24
Factor analysis
Simplifies relationships
16 of 24
Regression
Has 2 or more variables for prediction (instead of description). Two types: Simple regression, multiple regression
17 of 24
Assumptions of a regression
DV must be continuous and be interval or ratio. IV's must be nominal/ordinal/interval/ratio (if nominal must be 2 categories). Large enough sample size (20 per IV). Multicollinearity (excess of 0.90 massive issue)
18 of 24
T-tests
Comparison of two population means. One sample t-test (compares sample mean with known population or meaningful value), independent samples, paired samples (compares paired or matched samples or two means from repeated measures)
19 of 24
Wilcoxon t-test
Non-parametric test equivalent to t-test. Assumes DV is ordinal/continous level, IV is 2 categorical groups or matched pairs distribution of difference between groups is symmetrical
20 of 24
Mann-Whitney U Test
Compare differences between 2 independent groups when DV is ordinal/continuous but not normally distributed
21 of 24
Pearson's correlation
Parametric data, with continous data (IV and DV)
22 of 24
Spearman's correlation
Non-parametric, continuous data (IV and DV)
23 of 24
Multiple regression
Continuous data, with 2 or more IV's, parametric
24 of 24

Other cards in this set

Card 2

Front

Square root of variance and is a measure of dispersion

Back

Standard Deviation

Card 3

Front

Majority of measurements are 68%, which is +/- 1 SD

Back

Preview of the back of card 3

Card 4

Front

Produced using deviation from the mean= score/SD

Back

Preview of the back of card 4

Card 5

Front

Thinking there is a genuine effect in our population when there isn't. A probability alpha (0.5)

Back

Preview of the back of card 5
View more cards

Comments

No comments have yet been made

Similar Psychology resources:

See all Psychology resources »See all RDA resources »