Verbal reports and domain-specific knowledge: a comparison between collegial and retrospective verbalisation

Abstract

One way to investigate and account for the role of experience in dynamic decision-making tasks is to use a knowledge elicitation method, for example verbal protocols. Recently, methods for verbalisation by other subjects have been suggested as a way to elicit more information on thinking. One of these methods is here investigated in a quasi-experimental study on highly experienced and skilled train dispatchers. Collegial verbalisation is based on the procedure of videotaping practitioners while they perform their normal work tasks in their normal work setting. This is followed up by having a close colleague of the practitioner watch the video recordings and verbalise. The general hypothesis that verbal reports based on collegial verbalisations can provide protocol data that are close to the structure and content of verbal reports based on retrospective verbalisations was investigated. From this systematic comparison it is concluded that collegial verbalisations produce verbal protocols that are close to the retrospective verbal protocols on protocol and topic levels, but not on statement level. From this study, we conclude that the collegial protocols can be used as an independent source of data. It seems possible for a colleague to report verbally on a practitioners' observable behaviour in the same way as when the practitioner is doing a retrospective verbalisation and it may also be possible for a colleague to explain some of the non-observable behaviour of the practitioner.