SEEK Theory

How is knowledge important?

Like chunking theory as proposed by Chase and Simon (1973) is an example of a recognition-based paradigm, Holding’s SEEK theory falls under the more broad category of “knowledge-based” paradigms. So what is a knowledge-based paradigm?

Novice Physics Categorization
Figure A. Novice representation of a physics problem.
(Chi et al., 1982)

Knowledge-based theories

Knowledge-based theories stress the role of high-level, conceptual knowledge which, as Holding would argue, is logical given that chess experts (and experts across domains) rely on a large, well organized knowledge database. Knowledge-based theories draw extensively on the idea of levels-of-processing theory (Craik & Lockhart, 1972) which states that the level in which an object or stimulus is processed determines how well it has been encoded into memory and thus influencing recall. Stimuli that have been processed on a deeper level tend to be recalled with greater likelihood. This differential encoding leads to the idea that expertise isn’t simply about an expansive knowledge base (quantitative difference) but also that the knowledge base itself is organized differently for novices as opposed to experts (qualitative difference.)

Qualitatively Different Knowledge Base

Expert Physics Categorization
Figure B. Expert representation of a physics problem.
(Chi et al., 1982)

Chi et. al (1982) showed that experts organize physics problems on a more abstract level while novices seem to pay attention more to surface features thus leading to different knowledge representations. (See Figures A and B). Expert knowledge seems to be organized in a hierarchical manner (Freyhof, Gruber & Ziegler, 1992) The main evidence for these paradigms showing the role of knowledge in a top-down influence are expert systems (or computer-based models of these theories).

Some knowledge-based paradigms stress the idea that experts (like grandmasters in chess) recall a “corrected” version of a prototype, chunks that have been recategorized to obtain access to deeper semantic codes (Lane & Robertson, 1979) or to use high-level descriptors or verbal knowledge. One such theory is Dennis Holding’s SEEK (search, evaluation, knowledge) theory which argues explicitly against chunking theory (or purely recognition based paradigms). (Holding, 1992)

So what is SEEK theory?

Holding’s SEEK theory posits that experts are experts in general, due to their extensive knowledge base, and more specifically how they search and evaluate that items in the database. In essence, experts conduct more and better traces through their knowledge base (search), evaluate terminal positions in their search better (evaluation), and simply know more (knowledge). Because knowledge is organized in a more complex, and deep manner than in chunking, experts can store the “gist” of a position (without the need of rote memorization of layout like in chunking theory). Organization is therefore, more general and elucidates expert processing across domains. Masters can recall rapidly presented material based on “familiarity” with novel positions being classified along themes or as deviations from the prototype.

Yet, despite accounting for weaknesses of “chunking theory” in terms of accounting for forward thinking that seems to be displayed by experts, SEEK remains without a workable model and stresses verbal exposition (contrary to numerous findings of visuospatial advantages). (Gobet, 1998)

So just how does SEEK hold up to other theories of expertise?

SEEK Theory’s Explanatory Power

Valid XHTML 1.0! Valid CSS! Level Double-A conformance icon, 
          W3C-WAI Web Content Accessibility Guidelines 1.0 Bobby 508 Compliant! © Copyright Victor Long, Chandra Singh, and David Snitkof 2005
All Rights Reserved