An empirical investigation of the demographics of Top Management Team (TMT) and its influence in forecasting organizational outcome in international architecture, engineering and construction (AEC) Firms : a fuzzy set approach

Abstract

Whereas Top Management Teams (TMTs) are selected to fit a firm’s strategy, prior studies have evidenced that TMTs have significant impact on firm performance. The challenge of the two-way causality has been reflected in previous findings being ambiguous, inconsistent and sometimes conflicted. Pursing the same line of research may lead to incomplete and even error-prone conclusion. In contrast, this research suggests that inconsistency of findings among TMT demographics shown in prior work may point the possibility of studying the black-box nature of such relationships, and provide a tool to future forecast the organization outcome. More specifically, a multi-input (TMT demographics) multi-output (organization outcome) structure was used in this research to explore the future predictability power of TMT demographics for international Architects, Engineers and Construction firms (AEC firms). In order to build a reliable forecasting model, those contradictions were avoided by the utilization of artificial intelligence methods by training, testing and producing results without any prior assumptions or known structures. In particular, the Adaptive Neural Fuzzy Inference System (ANFIS) have been employed as a basis for constructing a set of fuzzy “if– then” rules with pre-tested input–output pairs. Three different forecasting strategies were constructed, the findings have demonstrated the learning and potential of the ANFIS model (time series based) in forecasting organization outcome, but at the same time, suggest that distinction should be established among different constructs of TMT demographics and outcome constructs. The results demonstrated that job-related demographics (i.e., TMT Educational Diversity, TMT Functional Diversity and TMT Tenure) could provide a satisfactory forecasting accuracy for the short-span (Liquidity) and medium-span (Cash Flow Stability and Capital Structure) outcome constructs. The future predictability power of other non-job demographics could not be evidenced in this research. Additionally, outcome constructs with dynamic nature could not be forecasted. Lastly, future research opportunities have been suggested for researchers. Most importantly, it includes the need to re-define diversity in the context of TMT composition (having different meaning as in: Variety, Separation and Disparity). Other methodological future opportunities are also suggested at the end of this study.

Date	8 Dec 2017
Original language	American English
Awarding Institution	École de technologie supérieure
Supervisor	Constantine J. Katsanis (Supervisor) & Sabah Alkas (Co-supervisor)