> ## Documentation Index
> Fetch the complete documentation index at: https://meta.niceshare.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Illusion of Control

> Illusion of Control is a personal chance set above the odds. Langer's 1975 office lotteries show how choice and extra days move the price and the trade.

<Info>
  **Category**: Effects<br />
  **Type**: Judgment bias<br />
  **Origin**: Ellen J. Langer, *Journal of Personality and Social Psychology* (1975)<br />
  **Also known as**: Illusory control
</Info>

<Note>
  **Quick Answer** — The **Illusion of Control** is an expectancy that your own chance of success is higher than the odds allow. Ellen J. Langer named it in 1975, after chance tasks were dressed with the cues of skill. Choosing a ticket, practicing a random path, or facing a nervous rival changed confidence. None of those cues changed the draw.
</Note>

## What is Illusion of Control?

The Illusion of Control is an expectancy of personal success that sits above the objective probability. The task is chance. A cue that belongs to skill — choice, familiarity, competition, or involvement — makes the chance feel steerable.

> An illusion of control was defined as an expectancy of a personal success probability inappropriately higher than the objective probability would warrant.

That sentence is Langer's definition in the *Journal of Personality and Social Psychology*, 1975. The miss is not a mood of certainty about a score you already earned. [Overconfidence Effect](/effects/overconfidence-effect) splits that later problem into a score, a rank, and a range. Here the question is whether a personal act is being treated as if it moved a draw that the act cannot move.

Blowing on a die before a family board game is the same shape. The breath is something a skilled throw might use. The faces of the die do not change. The Illusion of Control is what happens when the breath, the chosen number, or the careful setup is read as a better chance.

### Illusion of Control in 3 Depths

* **Beginner**: The cue is a chance task that looks like a skill. Before you raise a bet or keep a ticket, write the count of outcomes and the act that cannot change that count.
* **Practitioner**: Separate the piece you can steer from the piece you cannot. A checklist, a deadline, or a hedge is a real lever. A chosen raffle number, a practiced guess on a random path, and a nervous opponent in a card cut are cues.
* **Advanced**: Asking prices, refusals to trade, and ratings of control are not one number. A later review finds a medium average, and a wider gap for ratings of control than for the decision to act. A cent on a Yale card cut is not a trading loss.

## Origin

The name comes from one paper and six studies. The studies put skill cues onto tasks whose outcomes were chance.

Ellen J. Langer published "The Illusion of Control" in the *Journal of Personality and Social Psychology*, volume 32, number 2, pages 311–328 (1975). The byline is Yale University. The article is based on her doctoral dissertation in the Department of Psychology at Yale. Her thesis advisor was Robert Abelson. Reprint requests in the paper go to the Graduate Center of the City University of New York. The prediction was that competition, choice, familiarity, and involvement, brought in from skill situations, would make people inappropriately confident in chance situations.

The six studies follow that list. Male undergraduates at Yale cut cards against a confident or a nervous confederate. Office workers on Long Island bought a one-dollar football-card lottery, either choosing the card or receiving the card the previous buyer had chosen. A second lottery crossed choice with familiar letters or unfamiliar symbols. Telephone-company employees guessed which of three paths on a box would buzz, with or without a short practice, and with the stylus in their own hand or in the experimenter's. Visitors at Yonkers Raceway rated confidence in a track lottery at three times of night. A last lottery in a New York office building gave the three-digit ticket all at once or one digit a day.

> The objective contingency does not appear to be the crucial variable governing subjects' behavior.

That line is from the paper's implications. Langer's account is that people are motivated to master the environment, including events that look uncontrollable, and to avoid the discomfort of seeing no control. She ties a chronic sense of no control to passivity. [Learned Helplessness](/effects/learned-helplessness) is the page for stopping after uncontrollable harm, including when a later escape exists. This page does not carry those experiments.

## Key Points

Four cues keep the Illusion of Control tied to the task. The numbers below stay inside the study that produced them.

<Steps>
  <Step title="Write the count before the cue">
    A chance task has a count: cards in a deck, tickets in a pot, paths on a box. A skill cue sits on top of that count and does not rewrite it. Langer's list is competition, choice, familiarity, and involvement. Name the cue out loud before you treat a feeling as a change in the count.
  </Step>

  <Step title="A nervous rival is not a better card">
    In the first study, 36 male Yale undergraduates cut cards for a higher card, 18 against each confederate. Bets ran from 0 to 25 cents across four rounds. The mean bet was 11.04 cents against the confident confederate and 16.25 cents against the nervous one, t = 2.39, p \< .025. The first bet was 9.28 cents against 16.72 cents, t = 3.16, p \< .005. The competence rating was interpersonal, 4.8 against 3.17 on a 6-point scale, t = 5.46, p \< .005. It was not a rating of card skill.
  </Step>

  <Step title="A chosen ticket is still one ticket">
    In the second study, buyers who chose a football card asked $8.67 to sell it. Buyers who were handed a matched card asked $1.96, t = 4.33, p \< .005. Both groups already owned a one-dollar ticket, so the gap is choice, not ownership. [Endowment Effect](/effects/endowment-effect) is the page for valuing what you hold above an identical thing you do not hold. Langer also notes that a chosen card can matter for personal reasons. That alternative does not cover the studies that asked for confidence directly.
  </Step>

  <Step title="Extra days can hold a worse ticket">
    In the sixth study, people who received one digit a day were less willing to move into a lottery with a better chance. Of that group, 63.6 percent kept the original ticket, against 31.6 percent who had received the whole number at once, chi-square = 4.19, p \< .05. Mean confidence, on a 10-point scale, was 6.45 against 3.00, t = 4.16, p \< .005. The order of the digits was fixed. Thinking about the ticket did not add a number.
  </Step>
</Steps>

## Applications

Use the count and the cue whenever a draw, a raffle, or a random assignment is about to be treated as skill. Leave the lab figures in the lab.

<CardGroup cols={2}>
  <Card title="Write the count next to the act" icon="list-ol">
    Before a course raffle, a seat draw, or a family game, write how many outcomes exist and what you just did. Choosing a slip, blowing on a die, or watching a nervous opponent is the cue. The count is the chance. If the act is not on the count, do not raise the stake because the act felt skilled.
  </Card>

  <Card title="Keep a chosen number in the pot" icon="ticket">
    A chosen raffle number is one ticket in the pot you were told. In the letter lottery, the paper puts the original chance at 1 in 26 and the chance after joining the other pool at 1 in 21. The spoken offer said 26 buyers in the current lottery and 20 in the other. Choice and familiar letters each raised the share who kept the original ticket, Z = 1.83, p \< .05, one-tailed, for each main effect. The method says 13 men per cell. The table note says 15. If a later pool is larger for you, the earlier choice is a reason to notice the attachment. It is not a reason to refuse the count.
  </Card>

  <Card title="Score a rival on the task, not the coat" icon="user-group">
    A confident or awkward opponent changes a bet when the contest is skill. On the card cut, the mean bet was 16.25 cents against the nervous confederate and 11.04 cents against the confident one, in a sample of 36. In a hiring draw, a class lottery, or a coin for who presents, rate the procedure. A calm face is not a smaller chance for you. Those bets were cents. Each student was then set to win about \$2 on a later task regardless of the bets.
  </Card>

  <Card title="Separate the lever from the draw" icon="sliders">
    At work, a checklist, a limit, or a hedge is a lever. A forecast of a coin-like outcome is a draw. Fenton-O'Creevy, Nicholson, Soane, and Willman, in 2003, gave 107 traders in four organizations a computer task for this bias. The abstract reports a significant inverse association with managers' ratings and with total pay. That is an association in those firms. This page does not print the coefficient, and it does not turn the association into a market return.
  </Card>
</CardGroup>

## Case Study

The counted field case is Langer's sixth study. Adult office workers in one New York building were invited into a lottery by a male executive who worked in the building and did not know the hypotheses. A ticket cost \$1. One ticket was allowed. The whole pot went to the winner. Tickets were sold on Monday. The drawing was that Friday. Anyone who shared an office with more than three people was left out. People who shared an office were assigned to the same condition. The high-involvement group had 17 men and 5 women. The low-involvement group had 14 men and 5 women.

The high-involvement group was told the three-digit ticket would arrive in parts. The first digit came that day, the second the next day, and the third the day after. The order of the digits was fixed. The low-involvement group received the full three-digit ticket on the day of purchase.

On the day before the drawing, or on the morning of it, the seller returned. Demand had exceeded supply, he said, so he had started another lottery. That pot was fixed at $25, the ticket still cost $1, and the drawing was also Friday. Twenty-five people had bought into the original lottery, so that pot was \$25. Twenty people were in the other lottery. He offered to exchange the ticket and asked them not to mention it. The paper calls the exchange an objective increase in the chance of winning.

Fourteen of the 22 high-involvement workers kept the original ticket. Six of the 19 low-involvement workers did. Those shares are 63.6 percent and 31.6 percent, chi-square = 4.19, p \< .05. Eight high-involvement workers and 13 low-involvement workers said they would trade. Anyone who agreed to trade was visited again, apologized to, and told that the remaining tickets had been lost. The better draw was offered and then withdrawn. The measure is the decision, not a completed move.

That afternoon they rated the question "How sure are you that your ticket will be selected?" from 1, not sure at all, to 10, a pretty good chance. The high-involvement mean was 6.45. The low-involvement mean was 3.00, t = 4.16, p \< .005. Because the rating came after the trade decision, Langer also compared only the people who kept the ticket. Those means were 8.36 and 3.67, t = 4.15, p \< .005. The 3.67 is the mean of the six low-involvement keepers.

A separate questionnaire, given to Yale students in a library and to Prentice-Hall employees in New Jersey, asked what differed between the two ticket procedures. Thirty said involvement, 16 said fun, and 12 said complexity, chi-square = 9.24, p \< .01. Those respondents were not the lottery buyers.

The lesson is the cue. Three days of a fixed number made the original ticket harder to give up, and the confidence rating moved with it, including among people who kept the ticket either way. The offices were small, the assignment was by office, and the better lottery was never actually joined. A share of 63.6 percent in this building is not a rate of gambling, and it is not a loss in a market.

## Boundaries and Failure Modes

The Illusion of Control is the wrong name when a stated price is treated as a paid sale, when a racetrack rating is treated as the same people growing surer, and when one average effect size is treated as every measure.

The first boundary is the famous price. The $8.67 and the $1.96 are what 27 choosers and 26 non-choosers said a hypothetical buyer in the other office should be told, on the morning of the drawing. The sale was not carried out. Fifteen people first said they would not sell, 10 in the choice group and 5 in the other, p \< .10. That split is the paper's probability, and it is not a result at the .05 line. After prodding, those answers ran from $3 to the pot of $53, and only 3 people named the whole pot. Langer writes that a meaningful card, rather than confidence, could explain the choice studies. Four women refused the lottery. The six women who entered asked $1.33 on average, against $5.89 for the men, t = 2.14, p \< .05. Langer reads that gap as familiarity with football. It is six buyers, not a general difference.

The second boundary is the racetrack, which Study 6 was built to replace. At Yonkers Raceway, 409 people rated a lottery on a 10-point scale twenty minutes before the first, fifth, and ninth races. Male means were 3.14, 3.35, and 4.05. Female means were 4.10, 3.92, and 5.11. Time was significant, F(2, 403) = 4.09, p \< .05, and so was sex, F(1, 403) = 7.58, p \< .01. The rise from the first period to the third was F(1, 403) = 7.94, p \< .01. The female middle mean is lower than the first, so the cells are not a straight climb. The people at the three times were not the same people. Langer says the result may have been an artifact. The prize that week was a \$2,000 college scholarship. The public-address system was talking about the next week's Cadillac, and many visitors thought the Cadillac was the current prize. These ratings are not bets placed, and they are not the office means of 6.45 and 3.00.

The third boundary is export, including the neighbors and the later average. Stefan and David, *Journal of Applied Social Psychology* 43 (2013), pages 377–386, state in the abstract a weighted mean D of .62, with a 95 percent interval from .49 to .75. Stefan's thesis resume of that accepted review says the PsycINFO search, for 1996 through 2010, returned 635 items, of which 20 papers entered. Those papers held 34 studies and 84 measures. One effect per study, unweighted, was d = .79, SD = .47. The sample-size weighted mean was D = .62, VarD = .15. Direct ratings of control were D = .79 across 10 studies and 629 people. Behavioral measures were D = .54 across 8 studies and 613 people. Choice, as an inducement, was D = .47 across 3 studies and 292 people, interval .31 to .63. The sequence of outcomes was D = .89 across 4 studies, and the interval ran from .04 to 1.74. Do not read .62 as the card-cut cents, the office shares, or a trading loss. Fong and McCabe, *Proceedings of the National Academy of Sciences* 96 (1999), pages 10927–10932, found that more involvement led subjects to place less value, not more, on lottery tickets. One account they offer is that involvement made losses weigh more. [Loss Aversion](/effects/loss-aversion) is that weighing. This page does not print their prices. [Optimism Bias](/effects/optimism-bias) compares one's own future with other people's futures. [Self-Serving Bias](/effects/self-serving-bias) assigns credit after an outcome. Neither one is a skill cue sitting on a draw.

## Common Misconceptions

Three shortcuts travel with the name. A chosen ticket is treated as a better chance. A stated price is treated as money paid. One average is treated as every measure and every market.

<AccordionGroup>
  <Accordion title="Choosing the ticket changes the odds">
    In the football-card lottery, choice raised the asking price from $1.96 to $8.67. Both groups held one ticket in the same pot. In the digit lottery, the order of the three digits was fixed before anyone saw them. Keeping the ticket passed up a pool the paper calls an objective improvement, 14 of 22 against 6 of 19. The act changes the feeling and, in these studies, the price or the trade. It does not add a ticket to your side of the count.
  </Accordion>

  <Accordion title="The $8.67 was paid, and the card bets were dollars">
    The $8.67 and the $1.96 are answers to a hypothetical buyer. The sale was not carried out, and 15 people had to be prodded off a first refusal. The card-cut bets were cents, on a scale from 0 to 25 cents: 16.25 against the nervous confederate and 11.04 against the confident one. Each of those students was then arranged to win about \$2 on a jelly-bean estimate, regardless of the bets. A cent in that lab is not a dollar lost in a market.
  </Accordion>

  <Accordion title="One effect size covers every cue and every job">
    The weighted mean of .62 pools one effect from each study in a 1996-to-2010 review. Ratings of control sat higher, at .79, than behavioral measures, at .54. The sequence interval, .04 to 1.74, is too wide to treat .89 as a stable size. Fong and McCabe's involvement result ran in the other direction from Langer's. The trader association is 107 people in four organizations, reported as inverse with ratings and pay. None of these figures is a personal loss, a clinic rate, or proof that every skill cue works the same way.
  </Accordion>
</AccordionGroup>

## Related Concepts

The Illusion of Control sits next to a too-sure judgment, a rosy future, ownership, losses, giving up, and credit after the fact. The links are neighbors. They are not other names for a skill cue on a draw.

<CardGroup cols={3}>
  <Card title="Overconfidence Effect" icon="bullseye" href="/effects/overconfidence-effect">
    Confidence that runs ahead of a score, a rank, or a range. Langer's expectancy is about a personal chance on a chance task. The 2008 quiz does not move onto this page.
  </Card>

  <Card title="Optimism Bias" icon="sun" href="/effects/optimism-bias">
    A belief that one's own future compares well with other people's futures. A better personal chance on a lottery is a different comparison from a better life than one's peers.
  </Card>

  <Card title="Endowment Effect" icon="hand-holding-dollar" href="/effects/endowment-effect">
    Valuing an item you own above an identical item you do not own. In the football-card lottery both groups owned the ticket. The gap that remained was choice.
  </Card>

  <Card title="Loss Aversion" icon="scale-unbalanced" href="/effects/loss-aversion">
    Losses weighing more than matched gains. Fong and McCabe offer it as one way involvement might lower, rather than raise, the value of a ticket. Their prices are not printed here.
  </Card>

  <Card title="Learned Helplessness" icon="link-slash" href="/effects/learned-helplessness">
    Stopping after repeated uncontrollable harm, even when a later escape exists. Langer discusses the discomfort of seeing no control. The shock experiments stay on their own page.
  </Card>

  <Card title="Self-Serving Bias" icon="user-check" href="/effects/self-serving-bias">
    Taking credit for success and handing failure to the situation. That story starts after the outcome. The Illusion of Control is the expectancy before the draw.
  </Card>
</CardGroup>

## One-Line Takeaway

<Tip>Write the count of outcomes beside the act, and do not let a skill cue rewrite a draw it cannot move.</Tip>


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.