EXPERIMENT WORKSHOP
Notes by Asiah Capponi
WEEK 1
FOUR ELEMENTS OF TRUE EXPERIMENT
Example: obedience to authority, the Milgram Experiment
1. MANIPULATION
The researcher manipulates one variable by changing its value to create a set of two or
more treatment conditions (independent variable). Having two (or more) different
conditions of the ind var → control and experimental conditions
2. MEASUREMENT
A second variable is measured for a group of participants to obtain a set of scores in
each treatment condition (dependent variable)
3. COMPARISON
Only when there is a manipulation there are different groups measures that can be
compared. When there is a significant difference than the manipulation is considered to
have had a certain effect
4. CONTROL
All other variables are controlled to be sure that they do not influence the two variables
being examined (ex. Environment needs to be kept constant)
CORRELATION, CAUSATION AND EXTRANEOUS VARIABLES
Correlation does not imply causation!
Correlation: relationship between two variables BUT you do not know the direction of the effect
Causality: you know the direction of the relationship, there is a cause and an effect
So, what do you need to prove evidence for a causal relation? Use an experimental research
strategy
The different scores of the dep var are caused by the ind var
UNLESS there is an extraneous variable
, EXTRANEOUS VARIABLE= all variables beyond the independent and dependent variables
→ in a true experiment the goal is to control for all of them and prevent them to become
confounding variables
CONFOUNDING VARIABLES= when a variable influences the dep var and it varies
systematically with the ind var
Controlling for potential confounding variables:
● Control by RANDOMIZATION
Groups composed by people with individual differences→ it’s okay as long as the general
characteristics of the groups are very similar between groups
● Control by holding a variables constant
Groups are composed by people that have the same characteristics (ex. I chose only people
that are 30 and males)
Problem: might be hard to generalize to other groups
● Control by matching
Notes by Asiah Capponi
WEEK 1
FOUR ELEMENTS OF TRUE EXPERIMENT
Example: obedience to authority, the Milgram Experiment
1. MANIPULATION
The researcher manipulates one variable by changing its value to create a set of two or
more treatment conditions (independent variable). Having two (or more) different
conditions of the ind var → control and experimental conditions
2. MEASUREMENT
A second variable is measured for a group of participants to obtain a set of scores in
each treatment condition (dependent variable)
3. COMPARISON
Only when there is a manipulation there are different groups measures that can be
compared. When there is a significant difference than the manipulation is considered to
have had a certain effect
4. CONTROL
All other variables are controlled to be sure that they do not influence the two variables
being examined (ex. Environment needs to be kept constant)
CORRELATION, CAUSATION AND EXTRANEOUS VARIABLES
Correlation does not imply causation!
Correlation: relationship between two variables BUT you do not know the direction of the effect
Causality: you know the direction of the relationship, there is a cause and an effect
So, what do you need to prove evidence for a causal relation? Use an experimental research
strategy
The different scores of the dep var are caused by the ind var
UNLESS there is an extraneous variable
, EXTRANEOUS VARIABLE= all variables beyond the independent and dependent variables
→ in a true experiment the goal is to control for all of them and prevent them to become
confounding variables
CONFOUNDING VARIABLES= when a variable influences the dep var and it varies
systematically with the ind var
Controlling for potential confounding variables:
● Control by RANDOMIZATION
Groups composed by people with individual differences→ it’s okay as long as the general
characteristics of the groups are very similar between groups
● Control by holding a variables constant
Groups are composed by people that have the same characteristics (ex. I chose only people
that are 30 and males)
Problem: might be hard to generalize to other groups
● Control by matching
