KHA350 Lecture Notes - Lecture 4: General Linear Model, Multivariate Analysis Of Variance, John Tukey

51 views7 pages
Research methods week 4: Two Factor ANOVA
- If not statistically significant but a considerable magnitude of effect size, don’t
need to increase sample size or change the alpha level
oEffect size should be enough
oWe tend to overly worship p values
- One way ANOVA: only one independent variable or factor
oBut can have two or more levels
- Two way ANOVA: Two factor ANOVA
oFACTOR- INDEPENDENT VARIABLE- FACTOR CAN ALL BE
USED INTERCHANGEABLY i
oTwo IVs with one DV
oEACH iv CAN HAVE MULTIPLE levels
oUsually written as an a x b design
Where a and represent the number of levels in factors A and B
respectively
End of semester exam
oRequires an assessment of how the two variables interact to affect the
DV
How do the IV’s impact on the DV: When there are multiple
ones, we are interested in
Individual effect
How they interact with each other to have an impact on
the DV
- Example: Eysenck study
oExtension of One-way ANOVA added age as an IV
o5 x (by) 2 experimental design
oAge: between subjects variable
oCondition: look for phrases in the write up to work out whether
between or within subjects design
People were allocated to one of five conditions
Between subjects independent variable
oSummarize mean scores for each condition in a table
Cell means: means within the cells of the table
Marginal means: means in the margin of the table (average
score in row or column regardless of the condition or age they
are in)
Marginal means of age
Marginal means for condition
oMain effects and interactions:
2 IV’s
allows us to ask a number of questions about how these effect
the DV
1. Is there an overall effect of age on recall, regardless of
condition?
To test this, look at difference between old and you
regardless of condition
Testing marginal means
find more resources at oneclass.com
find more resources at oneclass.com
Unlock document

This preview shows pages 1-2 of the document.
Unlock all 7 pages and 3 million more documents.

Already have an account? Log in
The main effect of age: its effect ignoring the other
effects
Need to be able to find these marginal means in SPSS
output
2. Is there an over effect of learning condition on recall,
regardless of age?
Marginal means of condition
Main effect of condition
3. Is there an interaction between condition and age?
Is the effect of condition on recall the same for both you
and old participants?
Examine the cell means for each combination of age
and condition
This is referred to as the interaction between age and
condition
Each of the ten different cell means in compared
If significant interaction: the effect of IV a is not the same for
all IV levels of b
The effect for age on DV is not the same at each of the different
learning conditions
If non-significant: the effect of IV is consistent at all levels of
IV b
The effect is the same
“a significant interaction tells you whether consistent or not…’
- One Way ANOVA calculations:
oScore = grand mean + treatment error + residual error
oCompare the amount of variation due to the group they are in,
compared to the total error
- Two way ANOVA calculations:
oAdding more IVs just means adding further to the model
find more resources at oneclass.com
find more resources at oneclass.com
Unlock document

This preview shows pages 1-2 of the document.
Unlock all 7 pages and 3 million more documents.

Already have an account? Log in

Document Summary

If not statistically significant but a considerable magnitude of effect size, don"t need to increase sample size or change the alpha level: effect size should be enough, we tend to overly worship p values. One way anova: only one independent variable or factor: but can have two or more levels. Two way anova: two factor anova: factor- independent variable- factor can all be. Used interchangeably i: two ivs with one dv, each iv can have multiple levels, usually written as an a x b design. Where a and represent the number of levels in factors a and b respectively. End of semester exam: requires an assessment of how the two variables interact to affect the. How do the iv"s impact on the dv: when there are multiple ones, we are interested in. How they interact with each other to have an impact on. People were allocated to one of five conditions.

Get access

Grade+
$40 USD/m
Billed monthly
Grade+
Homework Help
Study Guides
Textbook Solutions
Class Notes
Textbook Notes
Booster Class
10 Verified Answers
Class+
$30 USD/m
Billed monthly
Class+
Homework Help
Study Guides
Textbook Solutions
Class Notes
Textbook Notes
Booster Class
7 Verified Answers

Related Documents

Related Questions