Authored by Dr. Baron Wolf
University of Kentucky
University of Notre Dame
When we make decisions throughout our day, whether personal at home or strategic in the workplace, we use data and information to assist us. This does not always mean we refer to complex charts, tables, or use regression analysis to see trends. Rather, we use our past experiences, intuition, and collected information to make the decision. Perhaps without knowing it, the decisions we make every day reflect a sound experimental design, not that dissimilar to the methods and design researchers throughout our organizations and universities are conducting daily in their laboratories.
Analytics can be defined as the “extensive use of data, statistical and quantitative analytics, explanatory and predictive models, and fact-based management decisions and actions” (Davenport and Harris, 2007, p. 7). However, the use of analytics in our decision making doesn’t have to be that complicated. And the use of data is everywhere. Take, for instance, your monthly home utility bills. If you pay your bills online through a website, you often land on a dashboard page that provides analytics related to your energy usage over the last billing cycle, trends related to the same time last month or the same time last year. In addition, even paper bills sent through the mail include brief analytics to tell the story of your utility use, past and present. Customers can use this data to inform their decisions on usage in the future. Having analytics allows individuals to see patterns more clearly and learn when something might be wrong. For instance, you might see a spike in your utility bill, so you look at the historical information provided and realize that this past month was unusually high, both in relation to the last three months and compared to the same time last year. You begin to put small pieces of data together, and ultimately you realize that something must be wrong with either the meter or the reading of that meter. You can use the data to make your case to the company for a reassessment instead of just paying the larger utility bill.
Analytics & Business Intelligence (BI) in the Research Enterprise
Analytics in research administration can provide critical insights: how we strategically align resources, support innovative ideas, and create efficient processes. Data gives us confidence in our strategies. Patil believes that “a data-driven organization acquires, processes, and leverages data in a timely fashion to create efficiencies, iterate on and develop new products, and navigate the competitive landscape” (2011, p. 2). The use of data within the university or research organizations is happening every day. However, having a firm understanding within the research administration profession on how analytics can drive growth and success is still in its infancy.
Just like a research scientist may conduct an experiment to collect data related to a substance’s impact on cancer cells, data using analytics and metrics can be used throughout the research administration profession to create growth strategies, impact efficiencies, and produce a better research enterprise. This can happen at multiple levels within the organization. Analytics can be a powerful tool for senior leaders across the organization in their efforts to make strategic funding and hiring decisions, identify areas to provide investment such as new buildings and create an overall growth strategy. Additionally, research administration offices can use analytics to impact operational efficiency, highlight workload imbalance, create innovative pathways to improve processes and help lead the expansion of the research enterprise.
Having an organized effort in the collection and organization of data and metrics is all you need to get started telling the story about the research enterprise. Graphically visualizing data allows decision makers to see trends, outliers, and areas for improvement. In addition, using information related to sponsored grants and contracts can help identify collaboration networks, increase productivity, and promote synergies within the academic community. For example, the use of social networking analysis provides a visual representation of key data points that can identify complex networks that exist. Likewise, network analytics can more clearly identify gaps or units working outside of the major network, that when combined can create synergy within the research enterprise. Below is an illustration of the use of network mapping to identify collaborations between academic departments working in a specific scientific area at a large research-intensive university (Figure 1). By visualizing the data, we can better align efforts to assist with research development opportunities by building team science, collaborations, and shared effort with experts across the institution.
Figure 1. Research Collaboration Network Analysis
BI Tools & Metrics for Research Administrators
Once you have the data, the visualizations can be done in Excel, Power BI, Tableau and other tools. One needn’t be a data scientist to do it, thanks to the online tutorials available via the websites of these tools and YouTube. A bit of time is all you need to create the visualizations and watch the data stories start to reveal themselves. Research administrators are typically well-versed in the individual data elements they collect related to sponsored projects. Using this information, with the help of BI tools, can move an organization from one that collects and stores data to one that creates a data-driven culture that helps lead the research enterprise. With the assistance of BI technologies and the use of advanced analytics, units will be able to not only report financials related to their research, but they will be able to explore and see trends and opportunities for a wealth of information.
Analytics for Senior Research Leadership
The reports requested from senior leadership about research activity really run the gamut from the basic questions: "How many proposals did we submit last month or year?" "How many awards did we receive?" to more complex analysis using a combination of data: "How many of our submissions involve multiple investigators and what are their departments/colleges?" "Are multiple investigator proposals more successful?" In addition, “senior leadership” is not limited to the central research office. Data requests come from department chairs facing review and tenure decisions, college deans distributing F&A return, as well as Vice Presidents, Provosts and Chancellors wanting to raise the institution’s profile nationally, globally, or both. Here is a table of often-requested reports:
In addition to research activity, leaders are interested in operational reports that answer such questions as:
- Is workload spread appropriately among the entire staff?
- Has the administrator’s workload changed significantly in the past year, or is it about the same?
- Is workload per month balanced, or are there times when staffer may need help due to volume?
- If this person leaves, which sponsors and sponsor types should I focus on in training a new employee?
- Are most of this staffer’s proposals high-dollar, or is there a balance across all proposal ranges?
- Does this staff member have expertise in any particular sponsor type? If so, will this limit future professional opportunities? Or is this the staffer who can best train others in a particular sponsor type?
- Is staff work varied enough to keep them engaged? Or is it overwhelming and may cause them to walk?
In addition to using data related specifically to sponsored projects and grant proposals, leadership at various levels can use other relevant data to make decisions that impact the research enterprise. For instance, institutions typically have a wealth of information related to faculty and research staff. Analytics using BI tools can be used to see patterns in productivity and alignment among faculty. Distribution of faculty effort and whether that effort is sponsored or non-sponsored is critical in making decisions related to research space allocation, release time, tenure and promotion, and a variety of other programs such as F&A fund distribution for special pilot grant projects. Having this type of institutional data easily accessible to leaders helps provide efficient use of time and enhances the decision-making process. Using analytics and BI tools help drive decisions to be strategic instead of only being based on a one-time ask. For example, at most institutions, there are a few individuals that are considered "rockstar researchers" with very large funding portfolios. We might also assume that funding is somewhat equally distributed among all faculty. However, when we use analytics and we begin to look at the “story” the data are telling us we find something else to be true. Figure 2. below provides awarded funding by a faculty member at a research-intensive public university. Of the 786 funded faculty PIs, 43 bring in 50% of all funding at the institution. This means that 5% of the faculty produce 50% of the funding. Leadership can use this information to invest in individuals in the middle to help get them over the line and create more successful faculty. Identifying where investments can make the difference can produce growth outcomes. By using analytics and BI tools, the story shown in Figure 2. can be interactive and easily replicated for leaders to explore their departments, colleges, and centers.
Figure 2. PI Funding Analysis
Conclusion & Survey Project
A lack of data is not the challenge. It’s making sense of it all. We began this article believing that research Administration Data Analytics is the foundation for building growth and innovation, but we need data to back that up.
Our goal with this article is to start a conversation on the types of analytics institutions currently use, who is responsible for collecting the needed data, tools used to reveal the stories within the data, whether institutions have a defined data strategy, and the value seen in sharing our data for greater overall insights.
We invite you to visit https://uky.az1.qualtrics.com/jfe/form/SV_5jyw9rsEs6Ax9pr to help us begin gathering data on the present situation. We will share the results in a future journal article. This study has been reviewed and approved by the IRB at the University of Kentucky. Finally, we hope this understanding of the current situation opens a door to discussing future goals and aspirations. Thanks in advance for your help.
Davenport, T.H. & Harris, J.G. (2007). Competing on analytics: The new science of winning. Boston: Harvard Business School Press.
Patil, D.J. (2011). Building data science teams. Sebastopol: O’Reilly Media.