Stepping into the Boardroom: A novel AI-enabled framework for recognising empirical manifestations of Collective Leadership from textual data

Anonymous

TL;DR: We present a novel Natural Language Processing AI-enabled task for detecting collective leadership applied to hospital board text data, including an evaluation and experimental approach using ablation analysis and human review.

Abstract: The concept of Collective Leadership (CL, broadly speaking leadership within groups) is difficult to define and detect empirically. A promising avenue for detecting CL focuses on discursive approaches based on group interaction and ‘turning points' in the discussion, where participants concur on the need for action. In the absence of a defined NLP task for the detection of CL, we present a novel AI-enabled pipeline applied to publicly available hospital board text data, requiring minimal annotation thanks to in-context learning. To our knowledge, this research is the first to combine NLP and leadership theories. After presenting a language model architecture, we propose an experimental approach using ablation analysis and posit an evaluation set-up including a ‘human in the loop' to aid acceptability by organisational research scholars and support the development of an annotated dataset.

Paper Type: short

Research Area: Computational Social Science and Cultural Analytics

Contribution Types: Model analysis & interpretability, Data resources, Data analysis

Languages Studied: English