Translational Science, Open Science, and Accelerating Practical Impact

Richard A. Guzzo, Workforce Sciences Institute, & Christopher M. Castille, Nicholls State University

Connecting science and practice, an essential of I-O psychology, is being aided by two complementary movements: translational science and open science. Although these twin movements influence how research is conducted, together the two have much to offer that is good for I-O, making for better science and strengthening links between science and practice. Further, for research conducted in organizational settings, there is a science-to-practice catalyst: The large organic databases now commonly found there can enhance both science and practice when principles of translational science and open science are adopted. With this entry into TIP’s Opening Up column on all things open science, we explain the background for our optimism and the ways we see these developments strengthening I-O’s science–practice bond. We begin by explaining what translational science is, then connect this movement to the broader open science movement, and clarify how these twin movements support.

What Is Translational Science?

Translational science arose as a response to the need in biomedical research to shorten the path from findings to action. Imagine white lab coated bench scientists prolifically publishing new findings that are immediately consumed by journal-reading peers, but findings with implications that remain, for far too long, out of reach of users such as health care providers and therapy developers. Historically, the translation of biomedical research discoveries into novel, drugs, devices, diagnostic tools, or therapies takes on average about 17 years (Pitzen et al., 2020). Navigating clinical testing and approval processes can slow the conversion of innovation to practice. However, translational science aims to reduce this time lag in several ways, including by supporting teams of scholars and using the science of team science (Pitzen et al., 2020). Although I-O psychology certainly is familiar with matters of linking science and practice, as with other disciplines, it too can experience long intervals between the appearance of new research findings in our journals and their applications in practice.

One part of the solution, perhaps not surprisingly, is better communication of findings to user audiences. But the much bigger part of the translational science solution lies in changing the research process itself—the way in which research gets done. To that end, it has developed principles, summarized in Table 1, designed to accelerate the movement of findings to practice. These are “big picture” principles to guide strategic research choices about such things as objectives, research team staffing, and project direction all with the shared objective of getting research results into action with speed. We invite
 I-O researchers to consider how the principles could influence their own work.

Table 1

Translational Science Principles*

Our high-level summary of how Table 1 applies to I-O research is encapsulated in three themes.

• First, go big! Target research at major theoretical and practical issues simultaneously and/or address several issues at once to command from multiple stakeholders an appreciation of the value and complexity of evidence-based practice. Such work aims at both expanding our understanding beyond its current limits while also focusing on practical matters. Well-known examples include the work of John Maynard Keynes, physicists who were part of the Manhattan Project, Louis Pasteur, and molecular biologists. Keynes wanted to understand and to improve how economies functioned. Physicists working on the Manhattan Project wished to understand and to harness nuclear fission. Molecular biologists want to understand and to alter genetic factors. Pasteur wished to understand and to control the process of fermentation. Indeed, to quote Pasteur’s famous dictum: “there is not pure science and applied science but only science and the applications of science” (see Stokes, 1997). “Go big” doesn’t mean addressing issues as grand as these, but it does call for research with the intention of delivering findings that advance science and practice simultaneously. As an example, consider “Project AI,” a consortium with a very ambitious effort to thoughtfully guide the application of AI in assessment practices (Oswald, 2024).
   
• Second, be interdisciplinary! Building off our previous example with the Manhattan Project, Robert Oppenheimer, director of the Project, initially used the following kind of organizational structure. There were five key divisions in the enterprise: theory (T), experimental physics (P), chemistry (C) (and later chemistry and metallurgy [CM]), ordinance and engineering (E), and administration (A) (Hughes, 2002). No doubt our I-O readership can probably spot a problem with this structure: silos quickly form! As such, Oppenheimer had to reorganize the Manhattan project to bust through silos that had formed and create a more interdisciplinary work arrangement (Hughes, 2002). This example reveals how important a close connection between theory and experimentation is for bringing about useful applications. Collaborating in all phases of a project is key, from planning through data collection and analysis, with members of other disciplines and with representatives of user groups to enhance research quality and acceptance of implications for practice. Returning to Project AI, it can be a valuable “think tank” ecosystem whereby I-O and OB scholars partner with professionals in applied statistics, computer science, philosophy, as well as industry partners that include test vendors, HR professionals, and other industry partners (see Oswald, 2024). Such an interdisciplinary collaboration allows a variety of issues to be considered simultaneously (e.g., accommodating for disabilities in AI assessments, revisiting validation strategies, placing guardrails on AI assessments) with a consideration for applied use (e.g., by testing vendors).
   
• Third, be free! Let the nature of the theoretical and practical issues as well as context determine research methods, rather than current methodological conventions or fads, while maintaining rigor and reproducibility for impact. Table 1’s principles were formulated specifically to up the pace at which research goes “from bench to bedside” in the vernacular of biomedical fields and early evaluations show favorable evidence of impact on research reports, theories of diseases, and pharmacologic solutions (e.g., Vogel et al., 2021). The spirit and principles of translational science have now transcended biomedicine and are influencing the sciences broadly. Returning again to Project AI, as AI is such a disruptive force that is garnering much attention, pooling our resources via interdisciplinary collaboration allows us to maintain relevance and have the support for being rigorous in addressing important questions.

The Role of Open Science of Translating Science Into Practice

Open science is a widely influential movement for elevating the quality of scientific research and the trustworthiness of its findings. The UNESCO Recommendation on Open Science defines open science broadly as encompassing different movements and practices aimed at making scientific knowledge openly available and accessible to anyone, reusable by any parties (e.g., methods), to make science a more collaborative enterprise. Such activities facilitate creating new knowledge and sharing information for the benefit of science and society (UNESCO, 2021). As such, open science can aid the process of turning research findings into practice.

Research and publications in I-O are slowly being influenced by open science. Although open science practices have been on the rise in the social sciences generally (Christensen et al., 2019), gaps exist between its ideals and what appears in the literature both in what our journals support and what scholars actually practice (Aguinis et al., 2020; Torka et al., 2023). Given the effort involved in implementing all open science practices in any single study (see Hostler, 2023), it is likely there will always be gaps. That said, we believe that there is value in each scholar finding their small win while they aim to go big (see Castille et al., 2022). This can be something as small as making sure any claims are reproducible in principle (e.g., getting the summary statistics right; see Murphy, 2021) or placing all of your analytical code online.

Open science ideals will not always align well with the nature of organizationally based research and organic data. For example, public research reports are unlikely to be able to fully disclose details or share data due to concerns for risk, privacy, intellectual property rights, and competitive advantages. However, open science qualities of disclosure, transparency, shared data, and collaboration often characterize, quite extensively, projects within the walls of an organization. That is, open science in these respects is “lived locally” and their benefits to the process of linking results to practice can be real. Also, in organizationally based research with multiple stakeholders who have a voice and numerous sources of available organic data, research plans can shift and new data introduced midstream, events that run counter to the fully preplanned research encouraged by open science. A challenge for I-O psychology, then, is to appreciate the elements of open science embedded in organizationally based research without letting the pursuit of open science ideals stand in the way of appreciating scientific advances made by research in organizations (Guzzo et al., 2022).

Organic Organizational Data as Catalyst

Organic organizational databases are a special kind of I-O researcher’s resource: “massive datasets capturing human behavior” (Zhang et al., 2022, p. 124). Many such databases are available in organizations. One that is a very rich source for I-O research is the HRIS database. It contains extensive facts about individual employees, jobs, performance, careers, work groups, reporting relationships, workplace characteristics, organization structure, management practices, and more. Other I-O research-relevant databases include those created by technologies that capture applicants, training and development activities, emails and other app-driven communications, calendaring, customer interactions, production processes, and sensor data (see Guzzo, 2022, for further details and implications). The term “organic” indicates that variables in these databases are the natural products of everyday technologies and procedures, and they stand apart from “designed” research variables which are those created for research by researchers and their instruments such as tests, questionnaires, games, and wearable sensors (Zhang et al., 2022).

When used on its own or in complement with designed data, organic organizational data can serve as a catalyst to the process of connecting research to practice by activating many of the features of translational science (TS) and open science (OS) that accelerate linking results to practice. It can be a catalyst in several ways.

• Organic organizational data are shared data that facilitates cross-boundary partnerships. Not all such data are accessible by everyone in an organization, of course, but there is a strong element of data democratization in organizations as evidenced by self-service access to databases to query, create reports, update facts, and do research. Further, many constituencies within an organization have interests in, and detailed knowledge of, available data and know where to go for them internally for use in research that serves business interests. Consequently, data sharing facilitates partnerships for research that cuts across internal organizational boundaries such as those of function, geography, and line of business. Shared data and partnerships are factors that benefit connecting research to practice. (OS, TS)
   
• Data access and use are interdisciplinary. Especially in larger organizations, individuals with training in research and data analysis from a variety of disciplines have access to databases. Additionally, organizations’ “people analytics” teams with research responsibilities often are intentionally staffed by researchers from diverse professional backgrounds. There is thus a baseline of resident interdisciplinarity that can be leveraged to support efficient transitions from findings to practice. (TS)
   
• Transparency with organic data is a requisite. Transparency engages stakeholders, fosters clarity and acceptance of findings, and helps stakeholders visualize connections between findings and their implications for practice. Researchers’ “analysis datasets” typically are data-wrangled extractions and combinations of one or more original raw organic databases, and the point at which the analysis dataset comes into existence is a good time for a data validation review involving researchers and an organization’s community of interested parties. Such review, which is a routine part of the first author’s experience when working with organic organizational data, double-checks data accuracy, builds trust in forthcoming findings, and prevents fraud, as happened when faked data were added to authentic data provided by an insurance company for a study of honesty. Incredibly, one of the study’s authors asserts that very few researchers who do field studies double check the data (Bartlett, 2024). This case illustrates what can happen when researchers “take the data and run” from an organization. Adopting open science and translational science principles means keeping organizations engaged. (OS, TS)
   
• Organic organizational data enables replication. Organic organizational datasets can be very large and contain many relevant variables. They can capture large sample sizes. They can contain information from multiple places and over extended periods of time. Characteristics like these make it possible to perform replications—across settings, samples, different operationalizations of constructs, and so forth—as part of a single study. Replication research is thought to be conducted undesirably infrequently in organizational research (Advancement of Replications in Management Research, 2024). Research reports from organizationally based studies using organic data that contain replications of core findings are an effective way of quickly building a literature of trustworthy, sustainable findings. (OS, TS)
   
• Organic organizational data helps research go big and bold. One part of “big and bold” in transformational science is delivering replicable research, as discussed above, and the other part is trying for major advances in knowledge and practice. Organic organizational data can help make this possible, in a few ways. They can inform a broad range of issues simultaneously, thus expanding the scope of topics addressed beyond that attainable with only researcher designed data because of the costs and intrusiveness of designed data collection compared to the low-cost, unobtrusive accumulation of organic data. Further, organic organizational data can be relied on to introduce researchable issues that are highly important to organizational practice, which when included in a research project that may originate with purely theoretical interests heightens the organization’s interest and stake. Anticipating and answering an organization’s “so what” questions is crucial to bridging findings and implications. (TS)

Conclusion

One of I-O psychology’s unique aspects is its dual emphasis on science and practice. The open science and translational science movements together help the field achieve those dual goals by changing how research is done. Hopefully, our call to embrace principles of these two movements is not regarded as yet another burden to “do more”—more sharing, more disclosing, managing research projects with more involved parties and such—that raises the demands on researchers’ time and skills. More is asked of researchers, for sure, but the benefits to a project’s scientific and practical impact are immediate and potentially large. This is especially so for research in organizations using organic data, where the opportunity exists to do more with more. Enhancing scientific quality and practical impact is the goal.

References

Aguinis, H., Banks, G. C., Rogelberg, S. G., & Cascio, W. F. (2020). Actionable recommendations for narrowing the science-practice gap in open science. Organizational Behavior and Human Decision Processes, 158, 27–35. https://doi.org/10.1016/j.obhdp.2020.02.007

Advancement of Replications in Management Research. (2024). Home page. https://www.arimweb.org/home

Bartlett, T. (2024). Is Dan Ariely telling the truth? Chronicle of Higher Education, Jan 18. https://0-www-chronicle-com.library.alliant.edu/article/is-dan-ariely-telling-the-truth?

Castille, C. M., Kreamer, L. M., Albritton, B. H., Banks, G. C., & Rogelberg, S. G. (2022). The open science challenge: Adopt one practice that enacts widely shared values. Journal of Business and Psychology, s10869-022-09806–2. https://doi.org/10.1007/s10869-022-09806-2

Christensen, G., Wang, Z., Paluck, E. L., Swanson, N., Birke, D. J., Miguel, E., & Littman, R. (2019). Open science practices are on the rise: The State of Social Science (3S) Survey [Preprint]. MetaArXiv. https://doi.org/10.31222/osf.io/5rksu

Guzzo, R. A. (2022). Organizational data and its implications for research and theory. In. K. R. Murphy (Ed.). Data, methods and theory in the organizational sciences. Taylor & Francis/Routledge, 3-27.

Guzzo, R. A., Schneider, B, & Nalbantian, H. R. (2022). Open science, closed doors: The perils and potential of open science for research-in-practice. Industrial and Organizational Psychology, 15, 495-415. doi:10.1017/iop.2022.61

Hostler, T. J. (2023). The invisible workload of open research. Journal of Trial and Error. https://doi.org/10.36850/mr5

Hughes, J. (2002). The Manhattan Project: Big science and the atom bomb. Icon Books.

Murphy, K. R. (2021). In praise of Table 1: The importance of making better use of descriptive statistics. Industrial and Organizational Psychology, 14(4), 461–477. https://doi.org/10.1017/iop.2021.90

National Center for Advancing Translational Science (2024). Translational science principles. https://ncats.nih.gov/about/about-translational-science/principles. Accessed 28 January 2024.

Oswald, F. (2024, March 2). Project AI: An agenda for AI-based employment testing. Personnel Human Resource Research Group Conference (PHRRG).

Pitzen, J. H., Dieter, H. L., Gronseth, D. L., Dahl, A. K., Boyle, V. L., Maran, T., Harper, C. M., & Gores, G. J. (2020). Transforming the practice of medicine through team science. Health Research Policy and Systems, 18(1), 104. https://doi.org/10.1186/s12961-020-00619-4

Stokes, D. E. (1997). Pasteur’s quadrant: Basic science and technological innovation. Brookings Institution Press.

Torka, A.-K., Mazei, J., Bosco, F. A., Cortina, J. M., Götz, M., Kepes, S., O’Boyle, E. H., & Hüffmeier, J. (2023). How well are open science practices implemented in industrial and organizational psychology and management? European Journal of Work and Organizational Psychology, 32(4), 461–475. https://doi.org/10.1080/1359432X.2023.2206571

UNESCO. (2021). UNESCO Recommendation on open science. https://www.unesco.org/en/open-science/about?TSPD_101_R0=080713870fab2000121b717d643702ce8f08c2926bcb10f3c1b903d661ea2c8057dcf4d2b48a6039085240ffdb1430007c8208ee037d1656f38a5707fe8e6e251e05b9acc85ebc54717033a1190b4786491fcd4408828b9ed3b24e0fc25cf0d1

Vogel, A. L., Knebel, A. R., Faupel-Badger, J. M., Portilla, L. M. & Simeonov, A. (2021). A systems approach to enable effective team science from the internal research program of the National Center for Advancing Translational Sciences. Journal of Clinical and Translational Science, 5(1), e163. DOI: https://doi.org/10.1017/cts.2021.811

Zhang, N., Xu, H., & Zhou, L. (2022). Organic data and the design of studies. In K. R. Murphy (Ed.). Data, methods and theory in the organizational sciences. Taylor & Francis/Routledge, 124-152.