Chapter 3
Outline
● Requirements of an Experiment
● Why Conduct Experiments?
Experiments, Critical Experiments,
and What-If Experiments
● Independent, Dependent, &
Controls
○ Null Effects
○ Floor & Ceiling Effects
● Interactions
● Between- vs. Within-Subjects
● Quasi-Experiments
Requirements of an Experiment
● What are the requirements to make a causal claim?
○ Covariance: Variables must be related to one another
○ Temporal Precedence: Does the IV come before the DV?
○ Internal Validity: Have we ruled out other explanations?
■ All three of these are the criteria to make a causal claim
● But Causal Claims Need to be Tested
○ How is this done?
■ An experiment!
“The behavior of elevator bystanders will cause conformity to such behavior.”
Was this an experiment?
No
● Manipulation
○ We must have an IV that is manipulated
○ This manipulation creates “levels” of our IV (e.g., control group vs. treatment
group).
○ Must have at least 2 levels
● Random Assignment
○ Participants must be randomly assigned to levels of the IV
● Control
○ We have control over the environment and the variables being measured
● Random Selection
○ Representative sampling
, ○ Individuals of the population all have an equal chance of being selected for the
sample
● Advantages
○ We have control!
■ The only thing that should be manipulated (i.e., changing for participants)
is the IV
■ This control allows us to attribute any changes in the DV to the
manipulation of the IV
○ We can make a causal statement!
■ Experiments ensure covariance, temporal precedence, and internal validity
○ Experiments are Economical
■ Naturalistic observation can be costly
■ Why?
● Observe participants in a natural environment, travel, coding,
training
● Disadvantages
○ Demand Characteristics → participants behavior changes based on perceived
expectations
■ Participants adhere to the behaviors that are demanded of them in an
experimental setting
○ Experimenter Effects → unintentional cues from researchers influence
participants
■ Experimenter tips off participant regarding what is expected of them
during the experiment
Why Conduct Experiments? Experiment, Critical Experiments, & What-If Experiments
● Why conduct experiments?
○ To test theories! → experiments allow psychologists to understand human
behavior
■ As psychologists we’re limited in learning about human behavior,
emotion, thoughts, etc.
○ Critical Experiments → compare competing theories (e.g., Waugh & Norman’s
study on memory)
■ A single experiment testing two theories
○ What-If Experiments
■ Conducting an experiment regardless of theory
■ Explore a new area that has little to no research
○ Replication → the more studies replicate a finding, the more confidence we
have in it
Outline
● Requirements of an Experiment
● Why Conduct Experiments?
Experiments, Critical Experiments,
and What-If Experiments
● Independent, Dependent, &
Controls
○ Null Effects
○ Floor & Ceiling Effects
● Interactions
● Between- vs. Within-Subjects
● Quasi-Experiments
Requirements of an Experiment
● What are the requirements to make a causal claim?
○ Covariance: Variables must be related to one another
○ Temporal Precedence: Does the IV come before the DV?
○ Internal Validity: Have we ruled out other explanations?
■ All three of these are the criteria to make a causal claim
● But Causal Claims Need to be Tested
○ How is this done?
■ An experiment!
“The behavior of elevator bystanders will cause conformity to such behavior.”
Was this an experiment?
No
● Manipulation
○ We must have an IV that is manipulated
○ This manipulation creates “levels” of our IV (e.g., control group vs. treatment
group).
○ Must have at least 2 levels
● Random Assignment
○ Participants must be randomly assigned to levels of the IV
● Control
○ We have control over the environment and the variables being measured
● Random Selection
○ Representative sampling
, ○ Individuals of the population all have an equal chance of being selected for the
sample
● Advantages
○ We have control!
■ The only thing that should be manipulated (i.e., changing for participants)
is the IV
■ This control allows us to attribute any changes in the DV to the
manipulation of the IV
○ We can make a causal statement!
■ Experiments ensure covariance, temporal precedence, and internal validity
○ Experiments are Economical
■ Naturalistic observation can be costly
■ Why?
● Observe participants in a natural environment, travel, coding,
training
● Disadvantages
○ Demand Characteristics → participants behavior changes based on perceived
expectations
■ Participants adhere to the behaviors that are demanded of them in an
experimental setting
○ Experimenter Effects → unintentional cues from researchers influence
participants
■ Experimenter tips off participant regarding what is expected of them
during the experiment
Why Conduct Experiments? Experiment, Critical Experiments, & What-If Experiments
● Why conduct experiments?
○ To test theories! → experiments allow psychologists to understand human
behavior
■ As psychologists we’re limited in learning about human behavior,
emotion, thoughts, etc.
○ Critical Experiments → compare competing theories (e.g., Waugh & Norman’s
study on memory)
■ A single experiment testing two theories
○ What-If Experiments
■ Conducting an experiment regardless of theory
■ Explore a new area that has little to no research
○ Replication → the more studies replicate a finding, the more confidence we
have in it
