Risk taking strongly influenced by sense of control
- Tuesday, November 20, 2001
CONTACT: Kim Osborne, (706) 583-0913, kosborne@...
WRITER: Allyson Mann, (706) 583-0914, tiny@...
RISK TAKING STRONGLY INFLUENCED BY SENSE OF CONTROL, SAYS UGA RESEARCHER
ATHENS, Ga. — A sense of control is a key factor in determining whether people
take risks or avoid them, says Adam Goodie, a University of Georgia assistant
professor of psychology. In a series of experiments, he found that participants
were more willing to take risks when they felt they could control the outcome
of a situation — even if they overestimated their likelihood of success.
Goodie presented the research Monday at the annual meeting of the Society
for Judgment and Decision Making, held in Orlando.
“People are really willing to take risks when they’re highly confident in their
own knowledge, even though it’s a bad idea,” says Goodie.
Goodie’s experiments build on a body of research indicating that people are
often overconfident about their knowledge. He asked participants to answer
trivia questions and rate their level of confidence for each answer.
Participants often specified a higher rate of confidence than was warranted;
when they estimated a 99 percent rate of confidence, their rate of answering
correctly was about 95 percent.
Goodie added a new element to the research: he used a point system to offer
participants bets based on their confidence level. With a high confidence
level, participants were offered a bet that gave a small reward (100 points)
for each correct answer but a large penalty (9,900 points) for each incorrect
answer. With a low confidence level, participants were offered what Goodie
calls “a more gentle bet”: a small reward (100 points) for each correct answer
and a slightly larger penalty (104 points) for each incorrect answer.
Over the long run, the riskier payoffs when confidence is high make them a
worse bet than the more gentle bets when confidence is low. Despite the odds,
Goodie found that participants were far more willing to risk a large penalty by
betting when they rated their confidence high. They were less willing to bet
when their confidence was low, even though the prospects for winning over the
long run were better.These results conflict with earlier findings that
indicated people rejected the odds of the high confidence bet when it was based
on a random event — the flipping of a coin, for example.
“When people are called to gamble on a random event, where there's a very large
probability of something small but good happening and a very low probability of
something big and bad happening, they don't want to do it,” Goodie says. “If
it's something they have control over, like their own knowledge of the world,
then they insist on doing it. That has significance for all sorts of decisions
we make in our lives.”
The most obvious example is the debate over airline travel since the events of
Sept. 11, Goodie says.
“People aren't afraid of driving, but they're afraid of flying even though air
travel is statistically much safer per mile traveled,” he says. “But people
feel safer if they're driving because they have control over the sources of
risk and uncertainty.”