Complex Experimental Methods Flashcards
interaction effect
effect of one independent variable depends on the level of another independent variable
IV1 depends on IV2
factorial design
testing more than one IV at a time allows us to test for interactions
- studies with two or more IV (also referred to as factors)
full factorial design
levels of IV 1 x levels of IV 2 = all possible combos
factorial notation
3 x 2 x 2
how many numbers are there? = # of IVs
what do each numbers represent = levels of IV
what does the product of the factors tell us? # of conditions
participant variables
IV that is non-manipulated
- variables are selected and measured abut not manipulated (e.g age)
IV 1
on cell phone vs not on phone
IV 2 (participant variable)
younger drivers vs older drivers
definition of main effect
is there an overall difference?
the overall effect of one IV on the dependent variable, averaging over the levels of the other IV
main effect calculations (table)
calculation
IV 1
level 1 level 2
IV 2 level 1 1 2 avg 1
level 2 3 4 avg 2
avg 1 avg 2
1. compare averages to see which IV has more of an effect on DV
2. subtract averages to see the difference between the IVs
how does IV 1 and IV 2 affect DV?
example answer: on average people who were alone (M = 9) reacted 3 minutes faster than people who were with strangers
difference in averages
defintion of interaction (table)
the effect of factor A on the DV depend on the level of factor B
- sign of results matter
how is IV 1 influenced by IV 2 ?
temp hot cold food hb 10 2 8 icesk 4 10 -6 6 -8 the effect of food is different based on temperature
the effect is temperature is different based on food
difference in differences
main effects in graphs
y
x x-axis: compare averages vertically y-axis: compare averages horizontally
questions to ask while figuring out main effect
- what is the IV?
- what are the levels of the IV
- what are the averages of each IV
- compare averages to determine main effect on DV
interactions for graphs
parallel = no interaction
non-parallel = interaction
interpreting interaction results
- compare endpoints vertially and horizontally then subtract
- the difference indicates difference in each interaction
same as what you would do in graphs
main effects for bar graphs
x - axis IV = compare middle
y - axis IV = compare height of bars with same color
interactions for bar graphs
method 1: do the differences between bar heights change as you go across x - axis?
method 2: connect the tops of the corresponding bars. would those lines be parallel?
not parallel = interaction
parallel = no interaction
between subjects factorial design
- all of the factors are manipulated between subjects
- each subject participants in just ONE condition