A framework for evaluating the criticality of software metrics: an analytic hierarchy process (AHP) approach

Purpose – Unless the key measures are identified and managed, the entire concept of measurement would end up making one step forward and two steps sideways. The key is to identify those measures that add value to the measurement approach per se. The present work is yet another contribution to further augment the purpose and spirit of the concept of measurement. Design/methodology/approach – The present research work strives to critically examine the relative importance of software metrics from the standpoint of their utility towards improving business performance. It illustrates how an absolute AHP framework can be formulated for the same. Findings – Criteria with respect to product, process and resource categories have been identified and decision alternatives have been explored in order to devise boundary conditions for classifying software metrics as critical, essential and redundant (CER). Practical implications – The new industrial era models and analogies are adequate to meet the rigours of the ever‐changing IT world. But, from the measurement and metrics perspective, it has long been a problem of plenty. The current study provides a methodology to streamline the cornucopia of measures into a consolidated definite set of metrics that are more meaningful and useful with respect to the set objectives. Originality/value – The lineage of measurement thinking is unswervingly attributing to a paradigm shift in the way metrics are evaluated and value creation is perceived and this has resulted in an approach that is more and more compatible with the Pareto principle. The current work adds fuel to such a thought.