task deliberately and generate a step-by-step solution, as shown by research on
physics problems (Anzai, 1991; Chi, Glaser, & Rees, 1982; Larkin, McDermott,
Simon, & Simon, 1980; Simon & Simon, 1978) and algebra-word problems
(Hinsley, Hayes, & Simon, 1977). Medical experts generate their diagnoses by
studying the symptoms (forward reasoning), whereas less experienced medical
students tend to check the correctness of a diagnosis by inspecting relevant
symptoms (backward reasoning) (Patel & Groen, 1991).
On the same theme, expert performers have a body of knowledge that not
only is more extensive than that for nonexperts but also is more accessible
(Feltovich, Johnson, Moller, & Swanson, 1984; Johnson et al., 1981; Voss, Greene,
Post, & Penner, 1983). Whenever knowledge is relevant, experts appear to ac-
cess it efficiently (Jeffries, Turner, Polson, & Atwood, 1981). The experts are
therefore able to notice inconsistencies rapidly, and thus inconsistent hypoth-
eses are rejected rapidly in favor of the correct diagnosis (Feltovich et al., 1984;
Johnson et al., 1981). On presentation, information in the problem is integrated
with the relevant domain knowledge (Patel & Groen, 1986, 1991).
Similar characteristics of expert performance are found across different
domains of expertise. The studies cited earlier suggest several important char-
acteristics that can be more effectively studied in relation to tasks particularly
designed to elicit them in a more controlled manner. We shall consider such
research shortly.
Extensive Case Studies of Single Subjects. In contrast to the group studies dis-
cussed earlier, in which small numbers of tasks were used to elicit the cognitive
processes of experts, we shall briefly consider two examples of research efforts
that have used detailed case studies in order to describe the cognitive processes
underlying superior performance.
The first example draws on several case studies of calendar calculations.
Calendar calculation is the rather astounding ability to name the day of the
week on which a given date falls. For example, when asked on what day of the
week August 5, 1934, fell, such a subject would be able to say, correctly, that it
was a Sunday. A major interest in this curious ability derives from the fact that
several individuals with this skill have been severely mentally retarded, and
little is known about how the ability emerged or was acquired. Analysis of this
performance is further complicated by the low intelligence of the subjects. On
the basis of a task analysis, where no knowledge about calculation can be
assumed for these mentally retarded subjects, one is led to assume that the
subjects must have memorized the information for all dates.
Investigators have examined a fairly large number of individuals for whom
the ability of calendar calculation has been substantiated (for reviews, see
Ericsson & Faivre, 1988, and Howe & Smith, 1988). Most calendar calculators
can demonstrate that ability for only a limited range of years. All such subjects
examined have been unable to explain how they know the correct answers.
Some investigators, however, have been able to assess mediating steps by ana-
lyzing these subjects’ mumblings prior to reporting an answer. Other inves-
tigators have been able to obtain informative retrospective reports on mediating
steps. The most reasonable conclusion seems to be that the detailed struc-
tures of these subjects’ processes differ from subject to subject and rely on a
Prospects and Limits of the Empirical Study of Expertise 535