The simple dependency ranking system - A novel method for defence capability prioritisation

Abstract

The prioritisation of capabilities is an important activity in Defence, as decision-makers grapple with shifting factors such as costs, schedules and tasking in the face of a finite budget. While a number of methods for prioritisation exist, they usually require large amounts of input data, often in the form of judgments, and thus can be computationally and resource intensive, particularly when accounting for multiple criteria, different stakeholder perspectives and interdependent issues. A novel approach has been developed, namely the Simple Dependency Ranking System (SDRS), which provides an alternative method of arriving at a prioritised list that also takes into consideration the most important dependencies between capabilities. The methodology of the SDRS was specifically designed to be simple and transparent. The approach essentially consists of creating two ranked lists: one of individual elements and one of pairs of elements (that correspond to element dependencies). By making assumptions with respect to the relative scoring of the ranked elements, both within each list and across each list, a final score for each element can be determined. The result is a priority ranked list that has taken the most important element linkages into account, according to the stakeholders. While initially originating from social choice theory, the SDRS is shown to be related to established methods in requirements prioritisation and network analysis. By virtue of its design principles, the method has additional advantages: it is relatively quick to perform, it requires minimal data in its operation, it has the potential to be scalable, and it can be performed by individuals or groups. In addition, the explicit articulation of element dependencies permits the generation of a partial systems diagram that should reflect the core connected clusters of the system. As it is not comprehensive, it is unlikely to be the only method used in managing a portfolio of projects such as the DCP. However, it may be useful as a quickly repeatable approach to identifying areas of similarity and difference and a framing technique for group discussion and decision-making. As well as introducing the method, this paper also reports on an initial experiment devised to explore and test the approach. Twelve volunteers took on roles as Defence domain stakeholders (Air, Maritime, Land and Joint) and prioritised a set of thirty-two Defence capabilities (eight from each domain). Two questions were of specific interest: • Does the process of identifying pairs improve the consistency of the final rankings amongst players? • Is it adequate to simply average individual responses, rather than attempt to achieve a group decision? The results of this limited experiment show some positive evidence for both questions i.e. some convergence towards an “agreed list” and the high similarity of averaged individual results with the team decision. While only preliminary, the initial results are encouraging and indicate that the method deserves further investigation and potential refinement.