Research on critical success factors (CSFs) in project management has produced highly divergent factor models, limiting cumulative knowledge and reducing the usefulness of fixed CSF lists for both researchers and practitioners. This paper aims to adopt a modelling perspective to examine whether such variability can be systematically explained by structural features of research design, rather than being attributed solely to contextual or sectoral differences.
A structured review of 81 peer-reviewed project management studies was conducted. Two researchers independently applied a standardised coding template. The number of CSFs reported in each article was modelled as a function of two interacting design dimensions: the underlying conception of project success (traditional, extended, mixed or sustainability-oriented) and the research methodology employed (quantitative, qualitative or mixed methods). Count-data regression techniques and robustness analyses were used to assess the individual and joint effects of these dimensions on CSF variability.
The results show that CSF variability is not random but largely driven by the interaction between success conception and methodology. More multidimensional conceptions of success are associated with systematically broader CSF sets, while sustainability-oriented conceptions produce extreme outcomes only under specific methodological configurations. Isolated effects of methodology are comparatively weak, highlighting the central role of interaction effects in shaping empirical CSF representations.
Rather than proposing another universal CSF list or a new theory of project success, this study contributes by modelling the structural drivers behind divergent CSF models. The study proposes an original and novel empirical lens through which CSFs are interpreted not as fixed constructs, but as adaptive and context-sensitive representations shaped by the conceptual–methodological configuration of the research from which they emerge.
