PSYB04H3 Lecture Notes - Lecture 11: Measuring Instrument, Internal Validity, Observer-Expectancy Effect
PSYB04
LEC 11
Chapter 12: Experiments with More Than One Independent Variable
Experiments with Two Independent Variables Can Show Interactions
●
Interaction effect →
whether the effect of the original independent variable depends on the level of
another independent variable
→ Intuitive Interactions
● Behaviours. Thoughts, motivations, and emotions usually involve interactions b/w two or more
influences
● Example: when ppl eat ice cream, they like their good cold more than hot; when ppl eat
p
a
n
c
a
k
e
s
,
t
h
ey like their food hot more than cold
● When psychological scientists think about behaviour, they may start with a simple link b/w an
independent and dependent variable
○ Often find they need second independent variable to tell the full story
→ Factorial Designs Study Two independent Variables
●
Factorial design →
one in which there are two or more independent variables (factors)
● In most common one, researchers cross the two independent variables → study each possible
combo of the variables
●
Cells →
unique conditions
●
Participant variable →
variable whose levels are selected (measured), not manipulated
○ Age, gender,and ethnicity
→ Factorial Designs Can Test Limits
● One reason researchers conduct studies w/ factorial designs is to test whether independent
variable affects different kinds of people; in different situations, the same way
find more resources at oneclass.com
find more resources at oneclass.com
● Form of external validity
○ Test independent variable in more than one group at once→ testing whether the effect
generalizes
○ Groups might respond differently to an independent variable
● Testing for moderators → using factorial design to test limits
○
Moderator →
variable that changes the relationship b/w two other variables
○ In factorial design → moderator = the effect of one independent variable depends on the
level of another independent variable
→ Factorial Designs Can Test Theories
● Goal of most experiments is to test hypotheses derived from theories
● Best way to study how variables interact → combine them in a factorial design and measure
whether the results are consistent with the theory
→ Interpreting Factorial Results: Main Effects and Interactions
● In analysis with two independent variables, there will be 3 results to inspect : two main effects and
one interaction effect
● Main effect
→
overall effect of one independent variable on the dependent variable, averaging
over the levels of the other independent variable
○ Simple difference
○
Marginal means →
arithmetic means for each level of an independent variable, averaging
over levels of other independent variable
○ Main effect = overall effect
● In factorial design w/ two dependent variables, first two results obtained are main effect for each
independent variable ; third result = interaction effect
○ Interaction effect = difference in differences
○ Can use a table to detect whether study shows interaction
○ Much easier to detect from a graph
■ If plotted as a line graph → check to see if the lines are parallel
● If lines are not parallel → probably is an interaction
● If lines parallel → probably is no interaction
○ Interactions are more important than main effects when a study shows both a main effect
and interaction
Factorial Variations
→ Independent-Groups Factorial Designs
● Both independent variables are studied as independent groups
● If 2X2 design → there are four diff groups of participants in the experiment
find more resources at oneclass.com
find more resources at oneclass.com
Document Summary
Chapter 12: experiments with more than one independent variable. Experiments with two independent variables can show interactions. Interaction effect whether the effect of the original independent variable depends on the level of another independent variable. Thoughts, motivations, and emotions usually involve interactions b/w two or more influences. Example: when ppl eat ice cream, they like their good cold more than hot; when ppl eat p a n c a k e s t h ey like their food hot more than cold. When psychological scientists think about behaviour, they may start with a simple link b/w an independent and dependent variable. Often find they need second independent variable to tell the full story. Factorial design one in which there are two or more independent variables (factors) In most common one, researchers cross the two independent variables study each possible combo of the variables. Participant variable variable whose levels are selected (measured), not manipulated.