Assess organizational capabilities with diversity recruiting through agent-based models
- 89 leaves : ill.
- Research studies project report ; no. SM-20- .
- Asian Institute of Technology. Research studies project report ; no. SM-20- .
A research submitted in partial fulfillment of the requirements for the degree of Master of Business Administration
Research Studies Project Report (M.BA.) - Asian Institute of Technology, 2020
Diversity of employment has been a concern on the management of human resources of organization. Many studies have been conducted including conscious and unconscious bias, cognitive and identity differences, and behavioral design. Several approaches have been applied to deal with this problem. A data-driven algorithmics is most cited one. Neuroscientists and software engineers have tried to come up with algorithm(s) that can minimize bias in recruitment to avoid possible homogenous workforce. However, even with the advent in machine leaning and artificial intelligent, it is still in the conceptual phase. This study explored another approach by adapting the simulation method called agent-based models. With the main concepts of procedural justice, choice architecture, nudging, and organizational capabilities, the agent-based models were developed. Several concepts and data, such as many-model thinking, philosophical profile of the individual, complex problem solving, and recruiting metrics were considered for computing the models. The analysis of data obtained from simulated models showed that a bias exits at the very first step of hiring, i.e., screening or interview. Recruiter with strong bias sensitivity prefers those who similar, either identically or cognitively, to them. It leads to the low hiring rate of different candidates. Contrarily, these recruiters are quite naïve when evaluating candidate of same type. However, with the process of group huddle selection, construct based on case-based design, and nudging to increase the tendency of individual with systemic discipline, the selection bias and groupthink were able to be minimized. Moreover, the result implies that a well-balance group of different individuals can be formed when the member adopts the ontological perspective. This leads to better performance, as evaluated by the percentage decrease on complexity of the problems.