Clarifying dynamics for organizational research and interventions: A diversity example

Abstract

Interventions backfire for many reasons, one being that the dynamics governing a system are not well-understood. To better explain organizational phenomena, and to intervene in ways that yield desired consequences, an appreciation of the core insights from dynamics may be necessary to include into a researcher's toolkit. Although substantial buzz surrounds the term dynamics in organizational science, conceptual missteps are present in the literature, ultimately limiting its application. We provide a coherent description of what dynamics encompasses by (1) advancing a definition and of dynamics and comparing it to other longitudinal concepts, (2) proposing theoretical principles to help researchers apply dynamics concepts to their own research, and (3) demonstrating how dynamics may refine our ability to explain organizational phenomena and devise practical implications. Plain Language Summary Organizational researchers and practitioners are faced with many complex issues and understanding how they unfold over time is difficult. Those difficulties often lead to unexpected results when we try to address those issues. Viewing organizations from a more dynamic lens can shed light on how phenomena evolve over time as governed by the iterative application of mathematical rules. Illumination of dynamics can assist with better targeting interventions and understanding their potential effects. A dynamics lens also emphasizes continuing needs to improve research methods, such as collecting longitudinal data and engaging in computational modeling. To illustrate these points, we discuss how diversity interventions in organizations may be better understood once dynamics are applied. For example, organizations wishing to improve the diversity of their work forces must focus both on why underrepresented groups enter the organization and why they might leave. Focus on only one or the other is unlikely to generate the desired effect. Once identifying the drivers of both entering and leaving and how those factors might reinforce each other, interventions to improve diversity might be better targeted.

A dynamic systems theory of intrateam conflict contagion.

“…These systems are often represented visually through a stock-and-flow diagram. In such diagrams, stocks represent the variables of interest and flows represent the processes by which the variables change levels over time (Olenick & Dishop, 2022). In our case, stocks refer to the state of an individual at a given point in time, and the flows are the mechanisms that move people between different states.…”

“…Second, our theory provides a dynamic systems representation that explicates the underlying mechanisms responsible for changes in contagion over time. That is, our theoretical representation speaks to how contagion can change over time as well as why contagion changes over time (Olenick & Dishop, 2022). As a result, our theory answers calls by previous conflict researchers (DeChurch et al, 2013; Korsgaard et al, 2008) who have advocated for increased attention to the multiple and dynamic processes surrounding conflict mechanisms (e.g., what combination of group processes best inhibits the spread of conflict?…”

A reply to commentaries on “Revisiting the design of selection systems in light of new findings regarding the validity of widely used predictors”

“…Regarding the first group of commentaries, we agree with Murphy (2023) that discussions of the validity of selection predictors should not be a "horse race." That is why we mentioned that composite validity and incremental validity are also important and that diversity (Olenick & Somaraju, 2023) and other criteria such as candidate experience (Jones & Cunningham, 2023) and cost/time (Harms et al, 2023) should be factored in. Similarly, we concur about the "validity for what" question (i.e., predictor-criterion construct congruence, Hough & Oswald, 2023) and the importance of careful selection design to increase the likelihood that the validity of a given predictor obtained in a specific situation (DeSimone & Fezzey, 2023) is at least as high as the metaanalytic average of Sackett et al (2022).…”