One-way Between Subjects ANOVA

Question 1

What are error bars?

Answer 1

Usually use standard error/mean (SEM)
SEM error bars aid inferential statistics
If 2 bars overlap the means most likely don’t differ significantly

Question 2

What is an ANOVA?

Answer 2

Used to analyse experiments with multiple independent variables with 2+ levels. Allows us to analyse complex designs and look at several IVs simultaneously as some questions not answerable using 1 IV

Question 3

What causes variability across ppts?

Answer 3

Individual differences
State of ppt (level of attention)
Measurement error
Can be within-group variability or residual variance

Question 4

What is partioning variability?

Answer 4

Quick and intuitive hand calculation of each source of variability (sum of squares)

Question 5

What are 3 types of variability?

Answer 5

Total mean square - all score variability
Model mean squares - variability explained by IV
Mean square residual/within-subjects - variability unexplained by model (within groups)

Question 6

How do you calculate the degrees of freedom?

Answer 6

Between groups = no groups - 1
Residual = no groups x (no people in groups - 1)

Question 7

How do you calculate mean square?

Answer 7

Divide sum of squares by corresponding df - expect MSmodel to be bigger than MSresidual

Question 8

What is the f ratio?

Answer 8

Variability that the model explains against variability model doesn’t explain

Question 9

What are effect sizes?

Answer 9

η2
0.01 = small
0.06 = medium
0.14 = large

Question 10

What is power?

Answer 10

What probability of detecting true effect size do you desire?

Question 11

What are post-hoc tests?

Answer 11

Strict criterion with reduced statistical power to control family-wise error rate
Different types e.g. Bonferroni, Turkey HSD