£

Clearly specify the problem:

What precisely is the decision to be made? What are the dimensions upon which the alternatives are to be evaluated? What are the key uncertainties involved? What simplifying assumptions are being made in representing the decision situation? How detailed should the representation be?

£

Source credibility and diversity:

It is a critical factor to attain trustworthy information from a variety of independent sources;

£

Assess consequences

(evaluative judgements):

How well do the dimensions represent the alternatives – do they include all relevant important features? In what way can we satisfactorily measure the dimensions? How should the different dimensions be weighted?

£

Assess uncertainties

(predictive judgements):

What information sources must be sampled to identify the degree of uncertainty involved – i.e., make probabilistic judgements – e.g., that there is a 75% probability that an event x will occur under certain specific conditions;

£

Evaluate alternatives:

What decision rule is to be applied – what is the specific basis upon which alternatives are to be compared? This involves weighing the assessed uncertainties with the assessed consequences;

£

Sensitivity analysis:

To what extent are the estimated consequences and uncertainties likely to be wrong? For example, how sensitive is the decision to changes in such things as using different weighting schemes for dimensions?

£

Gather further information?:

The decision will be sensitive to any lack of knowledge concerning certain variables, however, it should be asked: what are the costs and  benefits of securing additional information which might improve clarity and accuracy? Does it seem likely that any improvements would be outweighed by the amount of effort or time delays incurred to obtain this additional information?

£

Decision:

Having made the decision ask whether the analysis has been sufficient for the problem.