Understanding project managers' behaviour when using artificial intelligence for project control

Abstract

This research thesis contributes to Project Management (PM) literature and project managers' behaviour when using AI-based control analysis. The research premise is based on project managers using AI analysis as an early warning system for project control and studying their behaviour and decision-making when experiencing project escalation. For this purpose, the thesis adopts the theoretical framework of Behavioural Decision Theory, specifically the lens of Descriptive Decision Theory and the established concept of ”heuristics and biases”. Research on project managers’ behaviour during project control is well-established through the phenomenon of Escalation of Commitment(EoC). The main research question is answered through empirical research papers and is based on interviews with 22 individual project managers as research participants having experience using AI for project control analysis. The thesis contributes to synthesising the literature on AI in PM through a Systematic Literature Review and suggests two behavioural constructs “The Wait Effect” explaining EoC behaviour when using AI and “Override AI bias” explaining de-escalating behaviour. Further contributions include the variables influencing these constructs: “project uniqueness”, “explainable AI”, “dynamic project data”, and “information asymmetry”. The thesis further suggests the variable “trust in AI” can decrease biases on the use of AI for project control.

More information

Identifiers

ORCID for Nicholas Dacre: orcid.org/0000-0002-9667-9331

ORCID for Martin Kunc: orcid.org/0000-0002-3411-4052

Catalogue record

Export record