Hannah Samuelson is an Organizational Psychologist and Researcher at Fors Marsh Group. Her research areas have included the experiences of women leaders (including in the U.S. Congress), the intersection between personal values and leadership development, and gender-related issues in the military workplace. She has expertise in the use of computational modeling in the diversity, equity, and inclusion space and in the organizational sciences more broadly. Hannah holds a PhD in Social, Decision, and Organizational Sciences from the University of Maryland, College Park.
With inequalities systemic to the workplace being recognized now more than ever before, many leaders are realizing and acknowledging that their organizations may not be as diverse as they would like them to be. Members of the workforce, particularly millennials and Generation Z, emphasize diversity and inclusion as a critical consideration in their job searches—but organizations often fail to meet their expectations. Initiatives aimed at increasing diversity in the workplace not only have the potential to positively impact employees’ well-being and performance, but also to reduce social, health, and economic inequities more broadly.
Creating and maintaining a diverse, equitable, and inclusive workplace is a multifaceted effort that requires coordination from all levels in an organization. The complexity and uncertainty inherent to such efforts may be viewed as barriers to their implementation. However, the collection of rich data (e.g., data collected using Fors Marsh Group’s Inclusion, Diversity, and Equity Assessment (IDEA) Toolkit) combined with thoughtful, innovative analysis offers a meaningful avenue to reducing this complexity and uncertainty. Computational modeling is one method that can help organizational decision-makers minimize risks, understand the long-term impacts of initiatives enacted today, and ultimately, improve the efficacy of their efforts toward increasing diversity and fostering an inclusive climate.
What is Computational Modeling?
A computational model is a system, such as an organization, represented using equations and/or logical if–then statements written in a computer programming language, such as R. Simulations of a model, which involve allowing the relationships and processes represented within it to carry out over time from a starting state, can reveal the dynamics and emergent outcomes of those relationships and processes that may have otherwise been difficult to predict or measure, such as how small individual biases can become large societal rifts.
A particularly useful computational modeling approach for organizations wishing to better understand why their current diversity and inclusion metrics are or are not meeting expectations and how potential initiatives may impact these metrics over time is agent-based modeling (ABM). An ABM represents the behavior of individuals, teams, or even entire organizations (“agents”), the interactions between them (e.g., performance evaluations, networking), and their interactions with their environment (e.g., organizational policies). For example, researchers have built an ABM that shows that gender biases in both the hiring process and in the value of developmental opportunities can result in the persistent underrepresentation of women in senior leadership positions. Importantly, the ABM specifically reveals how these biases impact an organization’s demographic composition (by increasing women’s turnover due to experiences with tokenism, as well as creating a “sticky floor” effect, in which women remain “stuck” in the lower rungs of their organization due to decreased opportunities for promotion) and provides a platform for exploring what the reversal of that underrepresentation may look like.
Benefits of Computational Modeling
Simplified Data Collection
Computational modeling can help organizations circumvent common difficulties of collecting data on topics related to diversity and inclusion, such as identifying and including employees with concealable identities and reducing social desirability effects around sensitive topics. Furthermore, although many organizations already leverage surveys to better understand their employees’ opinions and behaviors, the organizations may not have the resources or time to administer surveys at the frequency required to capture the dynamics of a rapidly changing workplace. Research in organizational behavior can inform the processes built into a computational model that is further tailored to an organization’s needs, and data from a single survey administration can serve as the starting point in simulations. This approach can provide the data needed to guide organizational functions and policies with fewer requests for employees’ time and sensitive information.
Downstream effects of interventions and policies may be difficult to predict and may take years to manifest, potentially leaving organizations unprepared and resources misallocated. Computational modeling and simulation allow organizations to leverage the speed of modern computers to forecast the potential impacts of a diversity initiative or policy change. This decrease in temporal resources means leaders can gain more data points more quickly that can inform which initiatives and policy changes are likely to be successful and what risks may need to be mitigated. This is a powerful advantage when improvements to employee and organizational well-being are on the line.
The transparency of computational modeling is perhaps its most notable advantage for organizations that value openness and the inclusion of employees’ voices in decision-making. When designed to represent a specific organization, a computational model is an explicit, even if simplified, representation of that organization’s structures and processes. This clarity can better ensure that users can pinpoint the origin of risks and the negative outcomes from simulated organizational changes and, therefore, prepare to mitigate them. A computational model can also improve communication with stakeholders by providing a shared framework for discussions. Most importantly, when an individual’s perspective or lived experiences are not accurately reflected, the transparency and open-source nature of a computational model allows it to be iterated and innovated toward greater inclusivity.
Get Started with Computational Modeling
The benefits of computational modeling and simulation make it apt for gaining insight into the dynamics and potential outcomes of a wide range of workplace policies and functions (e.g., bias training, diverse innovation teams, formal and informal mentoring networks). For leaders wishing to challenge systemic biases in the operations of their organizations, to improve the working environment for all current and potential employees, and to positively impact their larger communities, computational modeling represents a critical instrument to add to their analytic toolkit. Fors Marsh Group’s organizational behavior research team can help you build the right tool for your organization’s diversity and inclusion needs. Contact an expert, here.