Expert opinion Gathering

When adequate past data about technology developments are not available and where the effect of a number of variables on technological development is not clear (the variables can be economic, political and social, and they can also be multi disciplinary interactions within the scientific community) the technique of expert opinion gathering, which is more of an intuitive method, can be quite effective. This technique is called the Delphi Technique. In this technique, a panel of experts, on a particular forecasting problem for technology development in the future, is formed. The names o the technical experts are not known to each other. A questionnaire to record the opinion of the experts is formulated and mailed to all the panel members regarding various quantitative and qualitative aspects and the interactions of a number of variables. It is observed that many panel members do not respond to the questionnaires and drop-out ate of the panel members may be quite considerable. It is recommended that for reliable results from Delphi, which is a multi round exercise, the number of panel members responding in the final round should be not less than 15. The first round of responses are pooled together, analyzed and a second questionnaire made out which is intended to gather more detailed or insightful information on the problem under consideration. For instance, if in the first round the panels members are asked to give their opinions regarding what breakthroughs in technology are probable in the next 20-30 years, in the second round one may ask them about the changes (probabilities) of some of these events occurring. The second round responses are again pooled together, analyzed and the quantitative statistics or qualitative opinion again presented to the panel members for their third round of opinions. In this round may present the statistics of the opinions gathered in the second round. The idea is to present different view points analytically so that the panelists can refine their earlier opinions, if found necessary to do so. In case the panelists have different opinions, they are asked to state reasons for their differing opinions. After the completion of the third, one may proceed to the next round where, again, the analyzed and processed opinions of the third round are fed to the panelists. There is really no restriction to the number of rounds administered but when the opinions more or less stabilize, one may be said to have reached the final round of the Delphi. The outcome of the final round will be the forecast with probabilistic quantitative and/or qualitative indications regarding the future.

Some precautions in administering the Delphi Technique:

1. The questions should be precise otherwise the answer may not be direct.
2. The Delphi expert panel members should be from various disciplines, so as to get a more integrated view of the future and to avoid any bias.
3. Questions should not be ambiguous.
4. The questions should be such as not to preclude any differing opinion. One should not close the door to a different point of view at the questioning stage.
5. The number of questions should be fairly limited and pertinent to specific areas of the respondents.
6. The Delphi technique which is an opinion-gathering and analyzing exercise in multiple rounds can be supplemented if necessary and if feasible by other methods of technology forecasting such as trend studies, regression analysts mathematical modeling and the like.

Dynamic Modeling:

Dynamic Modeling was made popular by J W Forrester. His book, World Dynamics inter-related many things such as internal resources, agriculture, population, pollution and industries. The dynamic models differ from the earlier given models. They can incorporate feedback information. There is always a relationship between input and output. Also, there is definite lag or lead time for the effect of one variable over another. These relations, the interactions of the different variables and the time relationship between them make the dynamic modeling process more advantageous than the static models.