Expert Knowledge
Grid has application wherever you want to acquire or understand expertise held by another - perhaps as the basis for an expert system design or part of a knowledge management system.
When the expert has a natural skill for a solving a problem or dealing with a situation, and you want to find out what that skill is - make that explicit - the repertory grid interview will gradually and gently push the expert toward putting verbal labels on pre-verbal concepts.
Whether you want to learn:-
- why a particular sales representative has a high success rate;
- how to differentiate employees who are effective in a new environment;
- how an interviewer is more accurate than others;
- how to design a process such as mental health differential diagnosis, using local medical expertise, as part of a medical expert system;
- how to build profiles to identify likely characteristics of serial offenders;
- how to understand knowledge familiar to a specialist;
- what special qualities characterise a successful school;
- what makes a particular detective more able to perceive which character is likely to be one of the bad guys;
- why one of your customer service representatives is so much better at handling complaints than the others - what does she perceive that is different?;
- and so on;
repertory grid provides a process for making expertise explicit.
Knowledge About Effective Performance
As an example, below is a partial list of bare constructs developed when a few people closely involved in implementing major changes in a retail bank got together to produce as many constructs as they could differentiating bank managers who were effective and ineffective in the new environment.
Their formal Purpose was:-
to explore the differences between successful and unsuccessful managers in the new environment
And the two against one comparisons were:-
- in terms of their performance at work;
- in terms of their performance on the assessment programme;
- in terms of their psychological test results;
- in terms of their personal history.
The constructs were:-
| good formal education | - | poor formal education |
| coped well with normal stress levels | - | went to pieces under stress |
| younger | - | older |
| truthful in psychological tests | - | lied in psychological tests |
| independent | - | group-dependent |
| distinguish accurately between what they can and cannot control | - | can't distinguish between what he can and cannot control |
| share information with their fellow-managers | - | doesn't share information with fellow-managers |
| manages conflict skilfully | - | gets into silly fights he can't win |
| ask questions because they're curious | - | ask questions because they're hostile |
| good understanding of the competitive environment | - | no understanding of the competitive environment |
| enjoy solving customer problems | - | dislike getting involved in customer problems |
| happy with new technology | - | afraid of new technology |
| set achievable but stretching goals | - | set goals which are too high or too low |
| take early action on problems | - | wait until problems become crises |
| proud of their staff when they achieve | - | hardly notices staff achievement |
| works long hours because he enjoys it | - | work long hours because of poor planning |
| better able to understand complex information | - | less able to understand complex information |
| understands his manager's goal | - | don't understand their manager's goals |
This is but a few of the nearly 90 constructs produced and prioritised in just 2 hours! With this wealth of information there is probably no need to complete a full grid analysis to look for highly correlated constructs.
Nevertheless, depending on the purpose, additional laddering, rating and differentiation, and the addition of ideal elements (e.g. the ideal manager and/or the manager guaranteed to fail) will produce much more information.
This is a very simple application of grid to design and administer. Additionally an expert is likely to find the process, which helps them articulate concepts which are important but to which they had not assigned verbal labels, highly satisfying as a means to enable them to share their expertise.
Four Related Resources
There are many examples of expert system development using repertory grid available by searching the web using the key words "expert system" and "repertory grid". Three such examples are:
- R. A. Stephens and J. G. Gammack Knowledge elicitation for systems practitioners: A constructivist application of the repertory grid technique Journal Systemic Practice and Action Research Issue Volume 7, Number 2 (April, 1994)
Abstract: This paper deals with the question of knowledge elicitation from a constructivist perspective. Repertory grids are examined as a means of exploring the perspectives of actors concerned with a problem situation. The issue of how to select a method to capture expert knowledge is dealt with and a practical application is provided of how the proposed approach might serve to inform the elicitor about the organization.
- Kenneth J. Preiss: A two-stage process for eliciting and prioritising critical knowledge Journal of Knowledge Management Dec 2000 Volume: 4 Issue: 4.
Abstract: An overview of knowledge elicitation, representation and prioritisation of key performance indicators (CSFs) necessary for success in direct selling in a dynamic market on a woman-to-woman basis is presented. The CSFs were elicited using the Repertory Grid (RepGrid) method. The RepGrid method allows for construct elicitation and the exploration of constructing indices, such as, frame differentiation, frame complexity, frame integration and construct centrality. The CSF constructs are then prioritised using the Analytic Hierarchic Process. The proposed two-stage methodology offers the benefit of eliciting individual mental constructs about critical knowledge, the representation of those constructs in a meaningful form, and the representation of repositories of knowledge of multiple entities across and within organisations.
- Boose/Bradshaw Knowledge Acquisition Reports from the Boose and Bradshaw knowledge acquisition research group at Boeing Computer Services
Knowledge Acquisition via Tracked Repertory Grids - Randy Wolf and Harry S. Delugach.
Abstract: One of the more valuable and flexible forms of knowledge acquisition is based upon the use of repertory grids. A useful extension of repertory grids can be created by providing a method of semantically linking associated constructs and repertory grids. This network of grids is a semantic network with nodes consisting of individual repertory grids and links acting as `tracks.' A track is a generalization of the laddering process used by repertory grid systems. These linked repertory grids which are acquired using the natural language interface of repertory grids can form an operational definition of a problem solving method.
Suggestions
Search for results from the Enquire Within site.
- Download a free evaluation copy of Enquire Within here
- Want to know more? Join the experts at the RepGrid User Group.
- Want to keep in touch? Join our mailing list.
- Twitter: Follow us at @EnquireWithin
- Email this page to a colleague.
- Send us comment on this page.
- Translate this page to your language? Try Google Toolbar
Contact Us
Telephone: Intnl + 64 4 384 5549.
Normal Office Hours 9:00 a.m. to 6:00 p.m. New Zealand Standard Time (9H00-18H00 GMT +12).
URL: http://www.EnquireWithin.co.nz
Mail:
Enquire Within Developments Ltd
94 Majoribanks Street
Mount Victoria
Wellington, 6011
New Zealand
Email us at: