The Rules of Asking
Research on people is bound by informed consent, minimized harm, the right to withdraw, and debriefing — the ethical frame that makes studies like Milgram's unrepeatable today. · 11 min
The last folio gave you a powerful tool: manipulate a variable, assign at random, and you can uncover a cause. But the people in a psychology study are not test tubes. They can be frightened, deceived, and hurt. So the tool comes with a second set of rules — not about what makes a study true, but about what makes it permitted. Those rules were written, in large part, because of one famous study.
Guess before you learn
Before running his obedience study, Stanley Milgram asked psychiatrists to predict: out of 100 ordinary people ordered to deliver a dangerous-looking 450-volt shock to a stranger, how many would go all the way? The experts guessed about 1 in 100. What did Milgram actually find?
About 65 percent delivered the maximum shock. No one was actually harmed — the 'victim' was an actor — but participants believed they were inflicting real pain, and many left deeply shaken. The finding was historic; so was the distress it caused. Keep your pencil mark: this study is why the rules in this lesson exist.
9–12
3–5
People are not like rocks or stars — they can be hurt or fooled. So studying them comes with rules: ask permission, keep them safe, let them quit, and tell them the truth when it is over.
6–8
Because the subjects of research are people, four rules protect them. Informed consent: they must know what the study involves and agree to it. Minimized harm: risks are kept as low as possible. The right to withdraw: they can leave at any point, for any reason, with no penalty. Debriefing: if any part was hidden, the researchers explain the real purpose afterward. A study that breaks these rules can be stopped, no matter how interesting its results.
9–12
Ethical research on people rests on four commitments. Informed consent means participants understand the procedure and its risks and agree without pressure. Minimized harm means the study is designed to keep physical and psychological risk as low as the question allows. The right to withdraw means anyone may stop at any moment without losing what they were promised. Debriefing means that when a study uses deception, its true aim is disclosed afterward.
These rules have teeth. Before a study runs, an ethics board reviews it and can refuse permission. This is why several famous studies could not be repeated today: their designs exposed participants to distress they had not truly agreed to, and no board would approve them now.
K–2
If you want to study how someone plays, you ask first. You tell them what will happen. They can say no. They can stop whenever they want. When you finish, you tell them the whole truth.
Undergrad
The modern framework crystallized in response to abuses — the Nuremberg trials, the Tuskegee syphilis study — and is codified in documents like the 1979 Belmont Report, which names three principles: respect for persons, beneficence, and justice. Informed consent operationalizes respect for persons; the harm–benefit calculus operationalizes beneficence; fair selection of participants operationalizes justice.
Deception complicates consent, since fully informing participants would spoil some designs. The compromise is bounded: deception is permitted only when the knowledge gained is important, no honest design would work, harm is minimal, and a thorough debriefing restores understanding afterward. An institutional review board, not the researcher, judges whether that bar is met.
Postgrad
Research ethics is best read as constrained optimization: maximize the epistemic value of a study subject to deontological side-constraints holding that some goods — a person's autonomy, their psychological and bodily integrity — may not be traded against knowledge, however valuable. Informed consent is the procedural expression of autonomy; it cannot be waived merely because a design is scientifically convenient.
The persistent tension is between validity and consent: naturalistic realism often requires that participants not know the true hypothesis, yet autonomy requires that they not be materially deceived about risk. Contemporary practice resolves this asymmetrically — deception about purpose may be tolerable, deception about risk is not — and locates the judgment in an independent board precisely because researchers are poorly placed to weigh their own study's worth.
informed consent
Agreement to take part given after being told what the study involves and what it might cost you. You cannot consent to a risk you were never told about.
The most unsettling part of Milgram's result was not the headline number but its shape. Before you read it, predict it. The participants were told to raise the shock one step at a time, past a wall of protests from the 'victim.' At what point do people start refusing? Sketch the share who are still obeying as the voltage climbs.
Why is this true?
Milgram's results were real and have been repeated. So why would no ethics board approve his study today?
Because the objection is ethical, not empirical. The study exposed participants to severe distress they had not genuinely consented to, and gave them no clear way to feel they could stop. A finding can be both scientifically valuable and ethically off-limits — and when they conflict, the protections win.
You now have both halves of the researcher's craft: the logic that makes a study true and the rules that make it permitted. That pairing carries through everything ahead. Next folio leaves method behind and turns to the machinery itself — the single cell whose firing underlies every thought, memory, and choice this course will study.
Practice — new ink and old, interleaved
1.Which of these is an empirical claim a psychologist could actually study?
2.Which statement could a psychology experiment test?
3.Without looking: which single rule does deceiving people about a study's risks always break?
Informed consent — you cannot agree to a risk you were never honestly told about.
How close were you? Grade yourself honestly — it sets your review date.
4.Name one thing the researcher must hold constant across both groups, and say in a sentence why it matters.
5.From the experiment folio: in a test of whether a drug improves memory, what does random assignment accomplish?
6.A study gives one group a caffeine pill and another a look-alike sugar pill, then times both on a puzzle. What is the dependent variable?
7.A researcher wants to secretly record students' conversations for a study. What does informed consent require?
8.A study briefly misleads volunteers about its purpose. In one sentence, say what the researchers owe them when it ends.
9.From the first folio: which of these is an empirical claim?