One-way ANOVA
Null and alt hypothesis
Ho: U1 = U2 = U3
H1: At least one groups is different
Null and alt hypothesis
Ho: U1 = U2 = U3
H1: At least one groups is different
Main effect and interaction effect
Main effect: The IV/factor affecting dv
Interaction: Both factors interact to affect the dv
What are Post hoc tests
Used after a significant difference was found in the ANOVA (p<a)
Effect size for all ANOVAS
Eta squared and etq partialed square
What is a factor? What is a level?
Factor: IV
Level: Conditions in the IV
How are repeated measures ANOVA similiar and different to one-way ANOVA?
Similar: 1 IV, 3+ levels
Different: repeated meaures
How many hypothesis?
Factor A
Factor B
Interaction
Why is it better to use ANOVA over many t-tests?
Reduces type 1 error (1 alpha)
Test wise alpha level
Alpha for an individual test
You can have 2 factors in a one-way ANOVA. True or false
false
3 advantages over the 1 way ANOVA
1.Less participants
2.Study change over time
3.Reduces individual differences
How many factors and levels
factors: at least 2 (just 2 in this course)
Levels: at least 2
What does this mean? How many groups of participants?
1 factor has 2 levels, and the other has 4 levels
8 groups of participants
Family wise alpha level
Cumulative error from many t-tests (a=.05 for each of the tests so if there is 3 tests a=~.15)
Assumptions for 1 way ANOVA
1.Independent observations
2.Normality
3.Homogenity of variance
Assumptions for repeated measures
1.Normality
2.sphericity
Draw a graph with interaction
lines will touch at some point
Assumptions for Factorial ANOVA
1.Independent observations
2.Normality
3.Homogenity of variance
A researcher studies going to bed (before 10pm or after 10pm) and showering (morning, midday, night) and their impact on hrs of REM sleep. What ANOVA is this?
Factorial
Between treatment SS= 10
Within treatment SS=
Total SS = 15
What is within treatment SS
5
15-10 = 5
What is total variance broken into
Within treatment and within subjects
Draw a graph showing no interaction
Parallel lines
What is a p-value?
Answers may vary
EX: -Chance of rejecting null hypothesis by chance alone
If partial eta squared is .14, what does this mean?
The partial eta squared 14% of the variance in shower time and bed time explains 14% of the variance in hrs of REM sleep.