DECISION-MAKING THEORY AND RESEARCH(including medical doctors and clinical psycholo-
gists) have found that linear models always do as
well as, if not better than, the clinical experts (see
Einhorn 1972). Similarly, bank loan officers asked
to judge which businesses will go bankrupt within
three years of opening was about 75 percent accu-
rate, but a statistical model was 82 percent accu-
rate (Libby 1976).
Arkes, Dawes, and Christensen (1986) demon-
strated this point with people knowledgeable about
baseball. Participants were asked to identify ‘‘which
of... three players won the MVP [most valuable
player] award for that year.’’ Each player’s season
statistics were provided. One of the three players
was from the World Series winning team, and
subjects were told that 70 percent of the time the
MVP came from the World Series winning team,
so if they were uncertain, they could use that
decision rule.
Participants moderately knowledgeable of base-
ball did better than the participants highly knowl-
edgeable, although the highly knowledgeable par-
ticipants were more confident. The moderately
knowledgeable group did better because they used
the decision rule more. Yet, neither group did as
well as they could have, if they had used the
decision rule for every judgment. How a little
knowledge can influence judgment will be further
explained in the next section on how people make
decisions.
DECISION PROCESSINGDecision theories changed because studies revealed
that people often do not make judgments that are
consistent with how the theories said they should
be making judgments. This section will describe
evidence about how people make judgments. Spe-
cifically, several heuristics will be discussed, which
are short cuts that people may use to process
information when making a judgment (Kahneman,
Slovic, and Tversky 1982, is a classic collection of
papers on this topic).
Availability. Consider the following questions:
What is the first digit that comes to mind?, What is
the first one-digit number that comes to mind?,
and What is the first digit, such as one, that comes
to mind? Kubovy (1977) found that the second
statement, which mentions ‘‘1’’ in passing, result-
ed in more ‘‘1’’ responses than either of the other
two statements. The explanation is that mention-
ing ‘‘1’’ made it more available in memory. People
are using availability, when they make a judgment
on the basis of what first comes to mind.Interestingly, the third statement, which men-
tions ‘‘1’’ in an explicit manner, resulted in fewer
‘‘1’’ responses than the first statement, which does
not mention ‘‘1’’ at all. Kubovy suggests there are
fewer ‘‘1’’ responses, because people have an ex-
planation of why they are thinking of ‘‘1’’ follow-
ing that statement (it mentions ‘‘1’’), so they do not
choose it, because ‘‘1’’ has not been thought of at
random. Thus, information must not only be avail-
able, but it must be perceived as relevant also.Mood. Information that is available and may
seem relevant to a judgment is a person’s present
mood (for a review of mood and judgment re-
search see Clore, Schwarz, and Conway 1994).
Schwarz and Clore (1983) suggest that people will
use their mood state in the judgment process, if it
seems relevant to that judgment. They contacted
people by phone on cloudy or sunny days, and
predicted that people would be happier on the
sunny days than on the cloudy days. That predic-
tion was verified. However, if people were first
asked what the weather was like, then there was no
difference in people’s happiness on sunny and
cloudy days. Schwarz and Clore suggested that
asking people about the weather gave them a
reason for their mood, so they did not use their
mood in making the happiness judgment, because
it did not seem relevant. Thus, according to Schwarz
and Clore, people will use their mood as a heuris-
tic for judgment, when the situation elicits actions
from people as if they ask themselves, how do I feel
about this?Quantity and Numerosity. An effect that is
also similar to availability was demonstrated in a
series of experiments by Josephs, Giesler, and
Silvera (1994). They found that subjects relied on
observable quantity information when making per-
sonal performance judgments. They called this
effect the quantity principle, because people seemed
to use the size or quantity of material available to
make their judgment. The experimenters had par-
ticipants proofread text that was either attached to
the source from which it came (e.g., a book or
journal) or not, which was an unfamiliar task for
the participants. They found that performance