The baby and the bathwater: research methods in construction management

This note is written in response to Seymour, D., Crook, D. and Rooke, J. (1997) Construction Management and Economics, 15 (1), 117-19. We argue against their narrow focus on the interpretative approach. Also, Seymour et al. are incorrect in implying that the 'rationalist approach' is necessarily quantitative. Our contention is that the choice of research approach in construction management depends on the nature of the problem. However, whatever choice of approach is adopted, it is important that the problem and associated key concepts are defined clearly and that the methods used, underlying assumptions and limitations are transparent and defensible. It is difficult to argue in favour of any single approach based purely on epistemological grounds as what constitutes knowledge is still an unsolved philosophical issue. Since construction management is a practical subject, we suggest that the choice of approach should be a pragmatic one: the approach that is likely to generate practical solutions should be adopted. Seymour et al.'s suggestion serves only to limit our choice of research tools. Furthermore, a lot of the research issues in construction management are practical problems which involve generalization of experience and formulation of hypothesis that can generate empirically testable implications. For problems of this nature, testability of hypothesis and reproducibility of results are important, and the naturalist approach (which is labelled 'rationalist paradigm' in Seymour et al.) of discovering causal relationship is more likely to produce general practical solutions. However, this does not deny the value of the interpretative approach, as it may be more suitable for certain types of problem. Moreover, in practice, an understanding of human behaviour 'from within' often provides useful insights for formulation of empirically testable hypotheses, despite the philosophical incompatibility of the interpretative and naturalist approaches.