Enacted Accountability, Information Integrity and AI-Supported Decision-Making in BIM-Enabled Projects: A Qualitative Codebook

Description

This dataset presents a qualitative codebook developed to examine enacted accountability, information integrity and AI-supported decision-making in BIM-enabled construction projects. It contains 115 codes organised into nine analytic domains, covering role boundaries and decision work, information integrity and readiness, AI-supported outputs, validation and professional judgement, trust and adoption, decision influence and pressure, accountability allocation, informal governance, and standards and governance readiness. The codebook was developed through abductive thematic analysis, combining concepts derived from the research framework with patterns identified in the interview material. Each code includes an operational description and an illustrative quotation to communicate its meaning and intended analytical application. The deposited codebook has been prepared for research dissemination in accordance with the consent agreement established with participants. Although individual codes may be supported by several quotations in the full analysis, only one quotation is presented for each code. Quotations were selected to minimise the possibility of recognising individual participants, organisations or projects while preserving the overall meaning and character of the analysis. The included quotations are therefore illustrative and should not be interpreted as representing the complete evidential basis for each code. The dataset is intended to support methodological transparency, scholarly interpretation and comparative qualitative research concerning accountability, information governance and AI-supported decision processes in BIM-enabled project environments. It does not contain the complete interview transcripts or disclose identifying information about participants or projects.

Steps to reproduce

Interview material was analysed using abductive thematic analysis. Data-near codes were developed alongside concepts from the research framework, compared across cases, refined for conceptual overlap and organised into nine analytic domains. Operational definitions were then prepared for each code. One minimally identifying quotation was selected to illustrate each code while maintaining participant and project confidentiality.

Institutions

• Nottingham Trent University

  England, Nottingham

Categories

Licence