What Is Behavioral Activation: The Simplest Depression Treatment That Works
Behavioral activation treats depression by reversing the withdrawal cycle through structured activity. Learn Lewinsohn's model, Jacobson's findings, and the evidence base.
Articles published in January 2025
Behavioral activation treats depression by reversing the withdrawal cycle through structured activity. Learn Lewinsohn's model, Jacobson's findings, and the evidence base.
In 1995, Claude Steele and Joshua Aronson gave Black and white Stanford students a difficult verbal test. When the test was described as diagnostic of intellectual ability, Black students performed significantly worse than white students matched on SAT scores. When described as a laboratory problem-solving task, the gap disappeared. Same students, same test, same items — only the social meaning of the test changed. Stereotype threat: the fear of confirming a negative stereotype impairs the performance it is feared to produce.
Deep learning uses neural networks with many layers to learn complex patterns from data, powering breakthroughs in image recognition, language, and more.
In 1913, Max Ringelmann had men pull a rope alone and in groups. Alone, each man pulled with about 63 kg of force. In a group of seven, each man's contribution dropped to 38 kg — 60% of their solo capacity. The loss could not be explained by tangled ropes. Ringelmann had discovered social loafing: the consistent reduction in individual effort that occurs when people work collectively rather than independently, measured across a century of research in tasks from rope-pulling to clapping to solving algebra problems.
In 1970, Henri Tajfel told Bristol schoolboys they preferred either Klee or Kandinsky paintings — a distinction Tajfel invented on the spot. He then had them allocate points to anonymous others. The boys consistently gave more points to members of their own arbitrarily assigned group. There was no history, no interaction, no competition. Just a label. Social identity theory explains how a meaningless categorization becomes the basis for discrimination — and why that tendency is not an aberration but a fundamental feature of human social cognition.
In 1971, Dennis Regan had a confederate give subjects a Coke during a break in an experiment. Later, the confederate asked subjects to buy raffle tickets. Subjects who had received the Coke bought twice as many tickets — even those who said they did not like the confederate. The gift had created an obligation independent of liking, independent of want, independent of the quality of the gift. Reciprocity: the social norm so powerful that Marcel Mauss called it the cement of civilization, and Robert Cialdini called it one of six universal weapons of influence.
In 1930, B.F. Skinner placed a rat in a box with a lever. When the rat pressed the lever, a food pellet dropped. The rat pressed more. When pressing the lever produced a mild electric shock, the rat pressed less. Skinner spent the next four decades mapping the relationship between consequences and behavior with a precision that transformed psychology, education, and animal training — and provoked a backlash that permanently changed how we understand learning.
Anthony Greenwald, Debbie McGhee, and Jordan Schwartz's 1998 paper introduced a test that could measure racial bias in milliseconds. You sort words into categories while a clock measures how long each sorting decision takes. When 'Black' and 'bad' share a response key, people who harbor implicit racial bias are slower than when 'Black' and 'good' share a key. The difference in milliseconds — your D-score — became the most debated measure in psychology. What it measures, whether it predicts discrimination, and whether changing it changes anything remain unresolved.
In the 1980s, Carol Dweck watched children in her Columbia lab respond to difficult problems. Some children said 'I love a challenge.' Others looked crushed and concluded they were not smart enough. The problems were identical. What differed was what each child believed difficulty meant — a signal to try harder, or a verdict on their intelligence. Growth mindset research: the implicit theory of ability that determines whether people use failure as information or as evidence against themselves.
In the 1960s, Aaron Beck was treating depressed patients using psychoanalysis — free association, dream interpretation, uncovering unconscious hostility. Then he started asking his patients what they were thinking during sessions. What he found was not repressed aggression. It was a stream of rapid, specific, self-defeating thoughts his patients barely noticed they were having: 'I'm stupid,' 'I'll fail,' 'No one likes me.' Beck had discovered automatic thoughts — and with them, an entirely different theory of what depression was and how to treat it.
In 1957, Harry Harlow placed infant rhesus monkeys with two wire surrogates — one that provided milk, one wrapped in terrycloth that provided comfort but no food. The monkeys clung to the terrycloth mother for most of the day and used her as a secure base to explore, running back when frightened. The wire mother — the food source — was visited only to feed. Harlow had disproved the drive-reduction theory of attachment: infants bond not to those who feed them, but to those who comfort them.
In 1989, municipal court judges were reminded of their own mortality and then asked to set bail for a prostitution case. The mortality-salient judges set bail at $455. The control judges set it at $50. A reminder of death had made people harsher, more defensive, more eager to punish those who violated their moral worldview. Terror management theory explains why: awareness of death drives human behavior in ways we almost never consciously recognize.
Women rate male job candidates more favorably than identical female candidates. Working-class voters oppose redistribution more strongly than the wealthy. Minority group members show implicit preferences for majority groups. System justification theory explains a pattern that confounds standard social psychology: people actively rationalize, defend, and legitimize the social arrangements that disadvantage them — because the alternative, believing the system is arbitrary or unjust, is psychologically more threatening.
In 1971, Edward Deci paid students to solve Soma puzzles they previously enjoyed — and found they spent less time on the puzzles during free time afterward. Paying people to do something they liked had made them like it less. Self-determination theory explains why: external rewards shift the perceived reason for an action from internal to external, and when the reward disappears, so does the motivation.
Artificial intelligence is technology that enables machines to perform tasks that normally require human intelligence, from recognizing images to writing text.
E. Tory Higgins showed children a cartoon animal that was either cheerful when it found its favorite food or sad when it didn't. Children encouraged to make the animal happy used eager, approach strategies. Children encouraged to prevent the animal from being sad used cautious, vigilant strategies — and performed identically well. Regulatory focus theory: the goal is the same, but the strategy for pursuing it divides the psychological world in two.
Jonathan Haidt asked subjects to evaluate a scenario: a family eats their dog after it dies in a car accident. No one is harmed. Everyone consents. Most people said it was wrong — but couldn't say why. When pressed, they invented reasons and then abandoned them when shown to be false, yet maintained the judgment. Moral Foundations Theory explains why: moral intuitions come first, and moral reasoning is constructed afterward to justify what we already feel.
After the Bay of Pigs disaster, John F. Kennedy asked his advisors: 'How could I have been so stupid?' The plan was transparently flawed. Every advisor in the room had doubts. None spoke. Irving Janis studied the Kennedy tapes and identified a pattern: the same dynamics that make groups feel unified and confident also systematically suppress the dissent, doubt, and critical analysis that good decisions require.
Mihaly Csikszentmihalyi interviewed chess players, rock climbers, surgeons, and composers and found they described their best experiences in nearly identical terms: total absorption, effortless action, loss of self-consciousness, distorted time, intrinsic reward. He called it flow. The puzzle his research revealed was disturbing: people experience flow more often at work than at leisure — yet consistently report preferring leisure. We have built a civilization that systematically produces the wrong kind of experience.
In 1969, Philip Zimbardo had NYU students administer electric shocks to another person. Half wore their normal clothes and name tags. Half wore hoods and oversized lab coats that concealed their identities. The hooded participants delivered shocks twice as long as the identified participants — to both a pleasant and an unpleasant confederate. Deindividuation: when identity is submerged, behavior changes in ways that cannot be explained by who the people are.
Nira Liberman and Yaacov Trope asked students to describe activities — taking a trip, eating breakfast, reading — either for tomorrow or for next year. For tomorrow, students mentioned concrete details: the sandwiches they'd pack, the alarm they'd set. For next year, they mentioned abstract purposes: adventure, health, learning. The activities were identical. Only psychological distance changed — and with it, how the mind represented everything about those activities.
Tversky and Kahneman's 1981 Asian Disease Problem: 72% of subjects chose certain survival of 200 people over a gamble for all 600. Reframed as deaths, 78% chose the gamble. Identical outcomes, opposite choices. Prospect theory — the 1979 Econometrica paper that replaced expected utility theory — explains why: people evaluate outcomes as gains and losses relative to a reference point, and losses loom twice as large as equivalent gains.
Stanley Milgram asked psychiatrists to predict how many Yale subjects would administer the maximum 450-volt shock to another person if ordered to by an experimenter. The consensus prediction: about 1-2%. The actual result: 65%. Milgram's 1963 obedience experiments remain the most disturbing findings in social psychology — revealing that the willingness to harm others under authority is not a feature of evil people but of ordinary social situations.
You have two $100 bills in your wallet: one earmarked for rent, one for entertainment. You spend the entertainment $100 on dinner. Later you find $100 on the street. Do you feel free to spend it on anything? Most people do — even though the money is identical. Mental accounting: the cognitive system of separate psychological budgets that makes rational fungibility impossible and shapes every financial decision we make.
Martin Seligman and Steven Maier gave dogs inescapable electric shocks in 1967. When later placed in a box where escape was easy, the dogs did not try — they lay down and accepted the shocks. Control dogs with escapable shocks learned to escape immediately. Learned helplessness: the experience of uncontrollable outcomes teaches organisms that their actions are futile — and that lesson transfers even when it is no longer true.
In 1999, Daniel Simons and Christopher Chabris asked subjects to count basketball passes in a video. A person in a gorilla suit walked through the scene for nine seconds, beat their chest, and walked off. Half of all subjects never saw it. Inattentional blindness: when attention is focused on a task, objects outside that focus — even conspicuous ones — become literally invisible. The science with consequences for aviation, medicine, driving, and security.
A neural network is an AI system inspired by the brain, built from layers of connected nodes that learn patterns from data to make predictions.
In 1998, Anthony Greenwald, Debbie McGhee, and Jordan Schwartz published the Implicit Association Test — a measure of automatic mental associations that operate below conscious awareness. White participants who explicitly reported low prejudice showed strong automatic associations between White faces and pleasant words. The science of implicit bias: what it is, how it is measured, what it predicts, and what the fierce debate over the IAT reveals about the limits of self-knowledge.
In 1984, Richard Petty and John Cacioppo told some students that a proposed exam policy would take effect at their university next year (high personal relevance) and others that it would take effect in ten years (low relevance). High-relevance students were persuaded only by strong arguments; low-relevance students were swayed by how many arguments there were and who was presenting them. The Elaboration Likelihood Model: persuasion takes two fundamentally different routes depending on how carefully people process a message.
A smoker who knows smoking causes cancer has a problem: the belief 'I smoke' conflicts with the belief 'smoking kills.' The discomfort of that conflict — cognitive dissonance — demands resolution. But resolution doesn't require quitting. It can mean deciding the evidence is overstated, adding the belief 'my grandfather smoked and lived to 90,' or concluding that dying young beats dying stressed. Festinger identified four escape routes. The science of how we make uncomfortable truths disappear.
Tversky and Kahneman spun a rigged wheel that landed on either 10 or 65, then asked subjects how many African countries are in the United Nations. Subjects who saw 65 guessed 45. Subjects who saw 10 guessed 25. The wheel had nothing to do with the question. Anchoring and adjustment: the cognitive process of starting from an initial value and adjusting insufficiently — leaving estimates biased toward wherever the starting point happened to be.