Forskning ved Københavns Universitet - Københavns Universitet


Duncan, K.D. and Praetorius, N. (1987). Knowledge Capture for Fault-Diagnostic Training. Advances in Man-Machine Systems Research, vol. 3, pp. 165-178. New Jersey: JAI Press.

Publikation: Bidrag til tidsskriftTidsskriftartikelForskningfagfællebedømt


Methods are described which have been employed to elicit verbally stated diagnostic heuristics from the utterances of process operators. Any attempt to access and represent expertise in verbal form will have to contend with (1) distortion of the representations by words, (2) distortions by serial production, and (3) efforts necessarily involved on the part of the expert. Moreover, theories of the acquisition of skill suggest that such difficulties are inevitable and can be anticipated and that any method or technique which overcomes them must therefore be a major research priority.
The development and subsequent use of diagnostic rules in training schemes together provide tests or criteria of how much knowledge has been effectively captured. The two tests consist essentially of (1) the coherence of operators' enquiry when they are "starved" of plant information, and (2) the intelligibility of strategic verbal statements to novice problem solvers
TidsskriftAdvances in Man-Machine Systems Research
Sider (fra-til)165-178
StatusUdgivet - 1987

ID: 224203878