The “Decision Quality” Framework consists of a process that defines a high quality decision. Its focus is on judging the quality of a decision at the moment the decision is made, rather than when the outcome is known. This separation of decision quality from its outcome has several advantages:
Framing aims to avoid the fallacy of choice, by ensuring that the right problem is being addressed. It involves ensuring that the scope and objectives for the decision are clearly defined, that the genuine “must-haves” are known (and distinguished from than the “nice-to-haves”). One of the challenges in this area is to be able to consider a sufficiently wide frame, whilst also working to narrow that appropriately and without losing important decision options.
This involves generating a full range of realistic and credible decision alternatives. The set of options should be quite diverse, at least initially, and the best features of each used or combined as far as possible. Disadvantages should be consider and overcome. Initially, “unrealistic” or non-feasible ideas can be included, as very often they may contain a valuable component or concept that can be included in (revised versions of) more feasible options. The final set of decision options should all be realistic, but still may represent a set of alternatives that are quite different to each other. Within each alternative, there may also be detailed variations or scenarios to consider.
Of course, modelling activity that is underway may also generate new ideas for decision options or variants that may have not been initially considered. However, in principle, modelling activity (and hence thee decision variants) is set within a decision frame, and cannot generally by itself identify the correct frame, context, objectives nor constraints within the decision situation: A process that is outside of the detailed model building an implementation is required, so that the ability to lead and conduct such a process should ideally form part of the skill set of a modeller.
This involves determining or anticipating the needs of multiple stakeholders, as well as how multiple criteria and objectives are to be valued against each other. The criteria can then be integrated within the model (as far as possible), and evaluated or traded-off in ways which are as objectives and rigorous as possible. One may need to take into account the nature of biases that may be present and the best way to mitigate their impact.
A quality decision required that the reasoning used is solid. For example, the choice of the modelling approach needs to be appropriate, and the logic needs to be sound. The analysis should be rigorous, structured and clear, and aim to generate insight. It may need to evaluate risks and uncertainties, so that decision-makers attitude to risk can be reflected in the decision process.
Clearly, the information provided to decision-makers should be relevant and accurate. This requires that input data or judgements are sound and accurate, and as free as biases as possible. It requires that one searches for data to ensure that all available information is used.
A quality decision requires a commitment to action. If – after a decision is made – there are delays in implementation due to unforeseen issues or objections, the the decision cannot be regarded as of the highest quality in principle. The framework seeks to create a commitment to action prior to the decision by anticipating and identifying potential implementation issues, objections, risks costs or complexity of execution. Even where such issues cannot be resolved beforehand, the principle that they will need to be dealt with can be established and accepted as part of making the final decision selection.