(Bereiter & Scardamalia, 1993; Brown, Collins, & Duguid, 1989). However,
this translation has not been particularly easy nor readily apparent (Stern-
berg, 2003). Unquestionably, many of the barriers to this translation have lit-
tle to do with the expertise research itself, but can be attributed to other so-
cial–political and educational circumstances beyond the scope of this chapter
(Alexander, Murphy, & Woods, 1996; Berliner & Biddle, 1995). Still, we must
look to the goals, premises, and methodologies of traditional expert–novice
research for plausible reasons for this research-into-practice conundrum.
Limitations of Traditional Expert–Novice Studies
Given its long and productive history, one might question the need for contin-
ued exploration of expertise or wonder about the value of pursuing alternative
conceptualizations. Despite the rich legacy of prior generations of expertise re-
search, limitations to those studies must be acknowledged, particularly if the
goal is to improve student learning and development in academic domains
(Ackerman, 2003). Five of those limitations are as follows:
- Highly selective problem-solving tasks and domains. The expertise re-
search has historically targeted a range of performance tasks and domains
that do not necessarily translate into learning in complex academic domains
(Alexander, 2003; Ericsson & Smith, 1991). Beyond the classic studies of
chess, there are investigations dealing with typewriting, table waiting, danc-
ing, as well as physics and medical diagnosis (Allard & Starkes, 1991; Erics-
son & Polson, 1988; Gentner, 1988; Patel & Groen, 1986). Within these prob-
lem-solving domains, experimental tasks have been very carefully selected for
their perceptual demands or their procedural characteristics (Holyoak, 1991).
Nonetheless, the correspondence that exists between expert waiters or typ-
ists and students learning history, mathematics, or other academic domain is
tenuous at best. Even when the problem-solving domains were more aca-
demic in nature, as with physics or mathematics, first-generation and second-
generation researchers did not look broadly at expertise within those do-
mains, but remained focused on how experts tackled particular kinds of
problems associated with those domains (e.g., Anzai & Yokoyama, 1984).
Thus, it remains questionable whether the highly consistent findings for ex-
pertise research derived from very purposefully chosen tasks, frequently per-
formed under more laboratory than real-life conditions, transfer to student
learning in the dynamic and messy context of the classroom. - Lack of developmental focus. Another limitation of past research on ex-
pertise has been its focus on extremes. In essence, traditional expertise re-
search has been a study in contrasts. Not only were the profiles of experts
built on their performance of very carefully selected problems, but also on the
performance of individuals at the other end of the expertise spectrum. The
278 ALEXANDER