Midterm 2 Flashcards
What’s in a factorial study design
At least 2 IVs with multiple levels
Each IV has a potential main effect
Main effect per IV
What’s a 2x2 factorial design
2 main effects with 2 levels
4 conditions in total in the study
How are main effects calculated
It’s not the observed data
Means are calculated by averaging the actual observations together
What’s an interaction
Looking for the difference in differences
Occurs when the effect of one IV depends on the level of another IV
Under what condition(s) does my IV affect my DV
How does the influence of one of my IVs on my DV differ depending on the influence of another IV?
How to interpret interactions
A significant interaction means you have found a difference in the differences
Each IV has a potential difference associated with it
With an interaction, the difference for an IV is not the same when you calculate it across the levels of another IV
What to look for in a graph
Error bars show variability across people, big bars are not good
Standard error should be about the same across conditions
They shouldn’t overlap
What’s a participant variable
Levels are selected (measured) but not manipulated
Not truly IV
Example: age
Can factorial design test limits
Yes
Does IV affect different kinds of people or people in different situations (generalizability of a causal variable)
External validity: test IV is more than one group at once
Show moderators: IV that changes the relationship between another IV and DV, moderator results in an interaction (effect of IV depends on the level of other IV)
Can factorial designs test theories
Yes
Are the results consistent with the theory
What’s a marginal mean
Means for each level of IV
Simple differences
Which effect is more important
The interaction effect is more important than main effects
How to describe an interaction effect
Describe with one level of the first IV and explain what’s happening with the second IV, do the same thing for the other IV
Expressions to describe an interaction effect (4)
It depends
Only for/Only when
Especially for
Can also look for participants variables (age, personality, gender, ethnicity)
What’s an independent-group factorial design
Between subject factorial
Both IV are studied as independent groups
2x2: 4 groups of participants
What’s a within-group factorial design
Repeated-measures factorial
Both IV are manipulated as within groups, requires fewer participant
2x2: 1 group for all 4 conditions
