Recently we attended the AIR Forum, the annual conference for institutional researchers. As a company, IData has a fair amount of direct experience with the practice of institutional research, especially around external and compliance reporting. In fact the genesis of the Data Cookbook is located in the problems institutional research has historically faced: no control over, and very little input into, the way data is collected and stored; definitional and reporting requirements that don't coincide with the operational and strategic reporting needs of the campus; a need to perform longitudinal and point-in-time analysis from fluid data systems and sets; insufficient knowledge of or access to (or both) campus data collections.
As a tool, the Data Cookbook, the data governance solution by IData Inc., allows institutions to document their data systems in some detail, all in one place; it allows for data definitions that includes business rules and processes around data collection and usage; it enables a robust data deliverables catalog that builds on these definitions and in doing so clearly distinguishes between operational, compliance, strategic, and analytical reporting and terminology; it supports efforts to publicly map and track data lineage and transformation.
Our interest today, however, is not primarily in the Data Cookbook, but rather the data management challenges and data governance opportunities facing institutional researchers (and analysts, data scientists, and other BI actors in general). Today’s post are some of my thoughts on the challenges that institutional researchers are up against.
For several years now, the number of sessions at AIR Forum related to data management and data governance has been growing, but this year the topic of data governance seemed to hit critical mass. Our presentation on getting started with data governance (click here to view recorded webinar on subject) was well attended, clients who presented on data governance reported similar attendance, and the conversations in and around the exhibit hall were lengthy and numerous. Data visualization, and of course the statistical analysis of data, continue to be hot topics at AIR Forum. Interestingly, the number of sessions around student learning outcomes assessment seemed lower this year (and very few of the conversations in the exhibit hall included this topic). Judging from this year's newest exhibitors, we wouldn't be surprised if over the next couple of years strategic planning and institutional accreditation become the hot topics--but that's another story.
The spate of presentations got us thinking: what is the most appropriate role for institutional researchers, or an institutional research (IR) or assessment office, in a data governance framework? If you think of data governance as the hub at the center of the wheel of data management activities, then you'd expect IR to be mainly interested in how it affects data warehousing, business intelligence, and analytics, and to be fairly indifferent to other aspects such as database operations, data architectures, even data quality, except to the extent that these aspects affect their work.
If you think of data governance primarily as the policies and practices that provide parameters and direction to data management, you’d have to wonder whether IR should be responsible for writing and promulgating policy, particularly in areas where it is not expert, or what level of involvement is appropriate. When it comes to data, IR is already a kind of administrator without portfolio, too often perceived as an auditor at best and police force at worst. Policies perceived as emanating from IR may be honored as much in the breach as the observance. And while policies and repeated procedures are important, my recommendation is that you come to them after you take many concrete and measurable steps.
If you think of data governance, like we do, as a pragmatic set of actions and instructions surrounding the collection, storage, use, sharing, and disposition of data, it's obvious that institutional researchers have expertise in some of these areas, and a vested interest in others. The analytical use of data, and the careful presentation of that analysis: those are institutional researchers’ stock in trade! Who better, then, to help develop and distribute governance activities in those areas in particular?
The practice of data governance that we recommend involves three key components:
- Identifying and empowering data stewards inside a working structure with clear and discrete responsibilities and areas of emphasis
- Creating and populating an accessible and current institutional knowledge base around data
- Facilitating a communication and collaboration process around all data users sharing and leveraging their knowledge.
Are institutional researchers data stewards? Other than when administering surveys, IR is not charged with much data collection, and its hands-on use of data mostly comes after significant processing and use. The challenge is how to provide guidance and support to that upstream work, which we know from personal experience, anecdote, and AIR's own publications to be an historic and ongoing problem at many institutions.
Is it reasonable to task institutional researchers with creating or overseeing a central knowledge repository? This comes closest to the kind of work we normally associate with IR. Still, a knowledge base will contain many elements that are distaff at best to the practice of IR, and it does not seem a good use of researchers’ time or labor for them to provide special oversight or coordination here. This is not to say they don't have much to contribute to a knowledge base, or that they wouldn't be heavy users of it as they seek to understand the sourcing, lineage, and daily usage of institutional data.
And while some institutional researchers may have abundant “soft” skills in the area of facilitating communication, managing the communications between and the collaboration among business offices is almost certainly outside their purview. Additionally, the coordination of data governance activities and processes is probably not the best fit for IR because its work in the service of decision support is more valuable than ever, and time spent elsewhere detracts from this work.
Having stipulated all these points, we recognize that data management personnel and data curatorial skills can be in short supply. While data governance can certainly be the focus of someone's job, it increasingly must also be a part of nearly everyone's. Institutional researchers, because they are often external to the business units that collect and work with operational data, can bring a strategic perspective and a cross-functional awareness to institutional data. Their feedback to data stewards, data consumers, and other data governance workers can be crucial, particularly when organized around analysis and decision support rather than quality or security. These are, of course, important areas of data management, and IR may well have the ability or inclination to help here, but we tend to think responsibility for these functions lies elsewhere.
The short answer to the question of what role institutional researchers should play in data governance turns out to be, wherever they are needed. But over the long term, as data governance becomes an integral part of the fabric and culture of institutions of higher education, we think IR can occupy a critical and strategic niche.
For pragmatic data governance, we see IR practicing data stewardship at the enterprise level, recognizing where competing definitions may be in play, identifying data quality issues in the aggregate rather than the record level, seeking access to data collections and using that access responsibly. Strategic planning requires IR’s input in the form of advanced analytics, longitudinal studies, and quantitative fluency – no organizational knowledge base is complete without some documentation of this kind of work, and the data required to perform it. So the knowledge base that supports your data governance regime must have an outsized role for IR. And whether IR acts as a conduit for data flows between offices and divisions, or as a clearinghouse of institutional information serving multiple constituencies, no IR office will achieve the success it would like without strong and consistent collaboration.
Ultimately, we’re largely agnostic about whether institutional researchers should occupy formal leadership roles in a college’s data governance efforts. That’s a decision that depends on many other questions such as:
- Is there support from management?
- Is there appropriate investment in technological and other resources?
- What is the focus of data governance: is it compliance? operational guidance? strategic decision making?
- Are the researchers in question comfortable with public roles and managerial responsibility?
What we’re not agnostic about, however, is that institutional researchers can and should provide leadership to data governance in action, perhaps by advising senior staff, perhaps by exemplifying good practices, perhaps by joining action teams that address specific issues, and almost certainly by patient advocacy.
Thanks for taking the time to review my thoughts regarding institutional researchers and their challenges. Feel free to contribute your thoughts.
IData has a solution, the Data Cookbook, that can aid the employees and the institution in its data governance, reporting, data-driven decision making and data quality initiatives. IData also has experts that can assist with data governance, reporting, integration and other technology services on an as needed basis. Feel free to contact us and let us know how we can assist.
(image credit StockSnap_OXWAMUDMLP_IRChallenge_BP #1068)