Heuristics and Biases in Audience Decision Making 223exposure to bodily harm. Attempts to promote healthy behaviors by providing consumers with
probabilistic information are generally unsuccessful.^156
Sample-Size Insensitivity: The Intuitive Appeal of Examples
A second bias related to the representativeness heuristic is insensitivity to sample size. Although
sample size is fundamental to statistics, consideration of sample size is rarely a part of the audience’s
intuitive decision-making process.^157 Advertisers’ statements such as “Four out of fi ve dentists
surveyed recommend sugarless gum for their patients who chew gum” persuade many consumers
despite the fact that the survey results are statistically meaningless without mention of the number
of dentists surveyed. Even when audiences receive information about a sample’s typicality, they will
often fail to use it.^158
Instead of basing their decisions on the appropriately sized sample, audiences tend to overgen-eralize from a single, easy-to-comprehend but unrepresentative example.^159 TV viewers often base
their perceptions of newsworthy events on a single example provided to them by the media.^160 The
person or event featured in the news story supposedly exemplifi es a whole category. Even when
viewers are presented with accurate information, frequent exposure to such exemplars can lead
them to make biased decisions.^161
A single example can have a stronger impact on audience decision making than statistics.^162 Forexample, mock jurors in a simulated capital sentencing hearing were more persuaded by a psy-
chologist who gave his personal clinical assessment of the dangerousness of the defendant than by a
psychologist who presented a statistically accurate actuarial assessment.^163 In a study that presented
the same information in a statistical and nonstatistical format, one group of high school teachers
read information about a new science curriculum that was in the form of a case study written by
a single teacher who had used the curriculum. The same information was given to another group
but in the form of a statistical summary of the fi ndings of 12 teachers who had used the curricu-
lum. The high school teachers found the case study to be much more persuasive than the statistical
summary.^164
Audiences are especially likely to disregard sample size when they encounter extreme exam-ples.^165 Audiences are also likely to disregard sample size when they encounter well-known
examples. For example, audiences weight identifi able victims more heavily than statistical vic-
tims.^166 As behavioral economists George Loewenstein and Jane Mather have observed, when
Rock Hudson and Magic Johnson were diagnosed with AIDS, the public’s concern for the
disease skyrocketed.^167
Surprisingly, experts, even expert statisticians, regularly fail to take sample size into account.^168A study of senior auditors planning an audit of a company’s internal control procedures fi nds that
auditors do not use the optimal sample size even when given all of the necessary information. Nei-
ther do they use the sample-size selection methodology the American Institute of Certifi ed Public
Accountants (AICPA) prescribes, nor do they notice sample-size errors.^169
The Causality Bias: The Appeal of Narratives and Stories
Audiences give the same information more weight when it is presented to them in chronological
order as an easy-to-comprehend narrative or story than when it is presented as a hard-to-comprehend
random list of facts. In a study comparing two versions of a travel brochure, one group of consum-
ers read a version of the brochure that described a vacation using a narrative form. Another group
read a version that described the vacation in a list form. Consumers who read the narrative version
evaluated the vacation much more positively than those who read the list of facts.^170