Using parameters to collect data
Parameter value assignment for a Knowledge Study using KSM is carried out to elicit the opinion of the knowledge domain expert regarding various factors that relate to the knowledge. Opinion is elicited as an integer value between 0 and 10. Descriptive terms are used to help the expert arrive at an informed opinion.
Having only one expert to provide parameter values could be seen as a limitation to the usefulness of the current KSM method and even the validity of the results derived from the values.
An alternative would be to have all of the experts involved, agree on every parameter value. Typically this would be 100 nodes x 6 values = 600 parameters on which to seek agreement.
The current method is not simply the provision of values by one expert; it is the provision of initial values by the expert that is focusing on a specific knowledge area followed by several opportunities to validate such values by the Knowledge Leader within the full context of the study.
The knowledge leader is a person with a good and overall knowledge of the whole area being studied. The knowledge leader will have also been told about his/her arbitration role concerning knowledge elements and parameter values.
A validation session is where the knowledge leader is shown the current state of the map and shown the current assignment of values and is asked to validate the information. It is the responsibility of the analyst/interviewer to bring anomalies to the attention of the knowledge leader. This is particularly relevant where map layout and learning dependency are to be questioned. In the case of parameter values, the knowledge leader will see the values assigned by various experts within the full context of the study. It is hoped that the knowledge leader can adjust values to remove bias and in certain cases to correct more obvious errors or misconceptions. It should be noted though, that misconceptions could be an important source of information for area managers and may not require ‘correction’ until viewed by the management team.
It is suggested that it is better to have initial parameter assignment by a focussed expert followed by several opportunities to validate values in context, by a more general expert with overall knowledge, than either of the individuals providing values alone. It is also better than two focussed experts discussing the values to arrive at a simple consensus. In previous work, a knowledge leader has sometimes disagreed with values and changed them and on other occasions, perhaps during the same validation session, has disagreed with values and made a point that the value provided is ‘interesting’ and left it alone.
There are clearly a range of options on how parameter values could be assigned. They could of course be assigned collectively after the map is complete, in a group meeting involving all experts. This may be a very useful activity but it will require a lot of staff time if it is to be done with care. The option should be available however; if an organisation feels that this method would lead to worthwhile outcomes.