Experimental Methodology August 31, 2004
Important Announcements • Office hours: Thursday 2-4 • Term paper due in class May 5th • Complete student survey
Objectives • Understand basic experimental methods • Systematically critique experimental studies. • Case example: The Stanford Prison Experiment • Break early to complete exercise for Tuesday’s lecture
Research Question: Bad Apples or the Barrel? • "A new Iraq will also need a humane, well-supervised prison system. Under the dictator, prisons like Abu Ghraib were symbols of death and torture. That same prison became a symbol of disgraceful conduct by a few American troops who dishonored our country and disregarded our values." [Bush Speech To The Army War College, 5/24/04]
How do we explain her sadistic behavior? • She is a particularly cruel woman • The prisoners are evil and therefore they deserved it • Someone “made” her do it • She was having a bad day • It was the desert heat • The army’s selection procedures favor sadistic people • The prison was poorly run
Stanford Prison Experiment: Study Background • Subjects: Undergraduate males at Stanford University • Subjects screened for pre-existing psychological problems/history of crime/drug abuse • Payment: $15 per day • Physical setting: Basement turned into mock prison with aid from consultant • Metal bars on all doors • Room with no windows used for solitary confinement
Random Assignment • Two Roles: Prisoner & Guard • Role determined randomly by flipping a coin.
Prisoners Picked up at home by armed police officers Finger-printed and placed in holding cell Searched, stripped naked and de-loused with a special spray Dressed in a smock with no underclothes Given ID numbers and were forbidden to use names Guards Given no special training Free to do whatever necessary to (1) Maintain law and order and (2) Command respect Guards dressed in identical uniforms Carried whistle and billy club Wore mirrored sunglasses to block eyes Independent Variable: Assigned Role
Prisoners Initially rebelled: Removed stocking caps, scratched off their numbers, barricaded themselves in cells 36 hours into the experiment: acute emotional disturbance, disorganized thinking, uncontrollable crying and rage. Planned two-week experiment called off after 6 days. Guards Forced prisoners to do pushups Sprayed prisoners with fire extinguishers Stripped prisoners naked Actively promoted aggression between inmates Forced prisoners to defecate in bucket instead of toilet Forced prisoners to clean toilets with bare hands Dependent Variable: Changes in Behavior
Lessons Learned • Was the aggression on the part of the prison guards due to their personalities or their situation? • What aspects of the design led to your conclusion? • How ordinary people can behave in extraordinary circumstances • How can prison abuse be prevented? • Was the experiment ethical? • Participants signed informed consent.
Basic Elements of an Experiment • Research Question: Driven by a process or a phenomenon • Random Assignment: To determine causality • Independent Variable: The experimental manipulation—in this case the role of prisoner or guard • Dependent Variable: Change in behavior caused by the independent variable (e.g. aggression)
Critiquing the Experiment MAJOR CRITIQUES: • Multi-faceted independent variable: What if the guards were not in uniform? What was the exact psychological mechanism? • Dependent variable not measured systematically: Which factor caused which behavior? How many guards behaved brutally and how many did not? • Demand characteristics: Principal investigator was also the “warden.” Did his presence influence the results?
"Say, what's the independent variable in this study?"(Professor Gordon Bower) “ To my surprise, I got really angry at him. Here I had a prison break on my hands. The security of my men and the stability of my prison was at stake, and now, I had to deal with this bleeding-heart, liberal, academic, effete dingdong who was concerned about the independent variable! It wasn't until much later that I realized how far into my prison role I was at that point -- that I was thinking like a prison superintendent rather than a research psychologist.” (Professor Philip Zimbardo)
Critiquing the experiment (continued) MINOR CRITIQUES: • Small sample size (only 24): Can limit the effectiveness of random assignment • Level of payment ($15) too much or too little? • Generalizability: (1) Convincing prison environment, but (2) Homogenous sample prevents us from generalizing to other populations (e.g. non-students, non-males)
Conclusion • A great deal of research in organizational behavior is either experimental or proceeds from assumptions derived from experimental work • Critically examine BOTH the flaws and the strengths of each study • Question the extent to which experimental research can be generalized to organizational settings