Interpretive structural modeling

A to Z of
Creativity Techniques
Previous
Improved Nominal Group Technique
Next
KJ-Method

Presume that you have a compilation of say 20 – 50 matters concerning some of the following:

  1. Issues
  2. Ideas
  3. Objectives
  4. Options
  5. SWOT elements (Strengths, Weaknesses, Opportunities, Threats) (see SWOT Analysis )

It is assumed these concerns have cropped up in a problem-solving activity, and you would like a group of people to reach an agreement with respect to ordering them in terms of some property such as:

  • Causality
  • Importance
  • Priority
  • Severity
  • Precedence

Alternative approaches to the above problem are Paired Comparison and Q-Sort. Interpretive structural modeling (ISM) (Warfield, 1982), is an adaptation of Paired Comparison, usually computer aided and capable of managing group input.

In theory each group member should be able to obtain every possible pair of items, and asked to state, which is prior (in terms of importance, severity, etc.). Each pair then ends up with a score showing what percentage of group member put them in a particular order, giving a comparative ordering of all the items. Computer support is usually required because the number of possible pairing goes up as the square of the number of items.

ISM has been used to assist local government bodies to prioritise large number of activities in circumstances of sever cutbacks. Activities were all listed, and then each decision-maker was asked to compare pairs of activity in rotation and to suggest which of each pair was the higher priority. A computer then summarised the large number of pair comparison it asked for to produce a network illustration that demonstrated all the activities that had been compared, linked by arrows that have the meaning ‘A has a minimally higher priority than B’.

When used for prioritisation, the objective is a single rank order, but ISM can also be employed to create networks, e.g. of causal relations. In this case, the comparison question might be: ‘Does A influence B or vice versa?’.