Towards a statistically validated model of perceived project complexity

Abstract

Project complexity is widely invoked to explain delivery difficulties, yet it is often used as a broad label rather than as a diagnostic construct that can be measured reliably and linked to organising responses. Existing frameworks propose multiple dimensions of complexity, but statistically validated survey measures remain limited, and there is continuing disagreement over whether emergent/dynamic complexity constitutes a distinct dimension or a temporal property of other forms of complexity.

This study provides an initial statistical validation of a survey-based measure of perceived project complexity derived from Maylor et al.’s (2013) Complexity Assessment Tool (CAT). We refined CAT-based items for survey use and collected responses from a large international sample of working project professionals recruited via Prolific. Using exploratory factor analysis followed by confirmatory factor analysis, we examined the underlying structure of the instrument, removed weak-performing items, and assessed the fit and reliability of the resulting measurement model.

Results support a four-factor structure that is interpretable in practitioner terms: (1) Organisation and Management, capturing the extent to which the project is set up for effective delivery (e.g., clarity, governance, resourcing, reporting, stakeholder alignment); (2) Institutional Environment, capturing the coordination demands associated with international dispersion (e.g., differences in location, language, time zone and currency); (3) Organisational Change, capturing the scale and significance of organisational and cultural change surrounding the project; and (4) Emergent, capturing anticipated change in project conditions over a six-month horizon. Notably, the Emergent items form a coherent factor rather than mirroring their current-state counterparts, providing empirical traction on the status of emergence within project complexity.

The study contributes a survey-based measurement model that both aligns with and recombines established complexity typologies in ways that reflect practitioner experience. It also strengthens the practical implications of complexity assessment by supporting a context-specific view of project organising: different configurations of complexity appear to call for different emphases in managerial discipline, coordination planning, change leadership, and adaptive response.