Across the country, thousands of decisions are being made every day within colleges and universities on topics ranging from admissions to graduation. Unfortunately, the information needed to inform these decisions is often not available quickly enough to be helpful.
We believe that the biggest challenge to establishing an analytics-driven culture within an institution is simply ensuring that data can be made reliably and consistently available to those who need it every single time that it’s needed. How to accomplish that, however, can get complex without the right strategy in place.
This is the challenge that Ivy Tech Community College—Indiana’s statewide community college and the largest singly accredited community college in the United States, educating nearly 175,000 students each year—endeavored to overcome and we can offer some suggestions from our experience.
1. Expect More from IT
First, are you fully leveraging your IT capabilities? It’s important to remember that, in every company, there is an enormous “data knowledge gap” between information technology and the functional teams. IT often knows everything there is to know about how to build, implement, augment, support and decommission various data systems while the knowledge of the data within those systems and the decisions being made from that data are usually held by employees throughout the rest of the organization. Because of this, most teams don’t think about including IT in their analytic endeavors.
If so, this is a huge missed opportunity. Data can seem like uncharted territory for many institutions and, while most functional teams will use what is already known about the data to drive new efforts, information technology can bring entirely new opportunities the table. Sometimes, you may even find ancillary solutions that are more impactful then what you set out to achieve!
For example, while working with administrators on an unrelated topic, one of our technical team members happened to observe a batch degree audit in progress and noticed that the output was a single PDF file for each student which then had to be manually reviewed by a staff member to confirm if the student had, in fact, completed the requirements for graduation.
This team member believed that the process could be greatly improved through data and worked to get the relevant information out of the degree audit system so that it could be more interactively used. Once complete, the administrators were able to do the same work in minutes that previously took days and had used the data to identify over 2,000 certificates, technical certificates, and associate’s degrees students who registered that term had earned and which could now be conferred.
2. Don’t Get Hung Up on Data Governance
Every organization needs some level of understanding and control around their data, but is an extensive data governance initiative involving dedicated resources and business data stewards absolutely necessary? The surprising answer is that, for many institutions, the answer is no.
When evaluating data governance strategies, consider the reasons behind the desire for data governance. What’s needed may simply be an ability to consistently and reliably know what the data means when it’s being used and that it always accurately reflects that meaning.
If this is the case, then you may be able to forego the more formal data governance practices and focus on simply ensuring that your institutional vocabulary is captured in one or more data dictionaries and that these definitions are appropriately reflected in your data.
3. Data Democracy
With an engaged collaboration between information technology and the functional teams along with an appropriately sized data governance effort, the next thing to decide is whether or not you’ll use a traditional centralized model or a decentralized model for data distribution. Most organizations use a centralized model to funnel requests for data through a specific team that has been formed to do so. For many, however, this can result in a somewhat Sisyphean effort as the increasing data demands of an active institution will, over time, result in slower and slower turnaround times. In these cases, data arrives too late for employees to make best use of it.
In contrast, a decentralized, or “data democracy” approach, can accelerate data use so long as the focus is kept on more than just providing access to data. To be successful, these teams must maintain a focus on the goal of empowering every employee with the ability to easily and intuitively use relevant data with little to no dependence on a centralized data team to meet their needs. Every decision made towards your analytics efforts must align with these goals—from integrating your data dictionaries and ensuring an uncomplicated user interface to outreach, training and support for your hundreds (thousands?) of new data users.
4. Engaging vs. Reacting
Often we hear from people that there just isn’t time to be engaged with all of the functional users within an institution. That the demands for data come so fast that teams struggle just to keep up much less have time to proactively engage with these areas. While a data democracy approach solves some of these problems and creates new bandwidth that can be used for outreach and training, consider creating the role of “Data Strategist” to more actively engage your organization around data.
One way to think of a data strategist is as an analyst who thinks like a business person, understands the vocabulary around data in information technology, and can effectively communicate the needs and challenges of both sides along with a deep hands-on ability to turn data into metrics and insights. The primary goal of a data strategist is help a designated functional area get the most out of your data capabilities.
5. Executive Sponsorship
While there may be individual leaders who champion the use of data for specific efforts, a college will never develop a data-driven culture so long as there is a willingness to accept anecdotal evidence and “gut feelings” as sufficient evidence to support action. Executive sponsorship that drives the use of data to support decision making is key to embedding this into the culture of an institution. Additionally, early successes in this approach will begin to quickly demonstrate the value of data-informed decision making across all functional areas.
Ivy Tech recently conducted an applied predictive analytics effort called “Project Early Success” that was supported by our president and involved calling thousands of students during the first few weeks of the term. These were students who we believed, based on our predictive analytics, may fail given courses based on specific behavioral attributes. Callers were provided with scripts that allowed them to provide targeted information to each student based on the reasons we believed they might fail.
About halfway through the term, there was some disappointment amongst the project team as it was “felt” that there were still a high number of non-passing grades being reported at midterm. In fact, the data showed just the opposite. In fact, the college experienced the largest drop in the number of midterm D and F grades as a percentage of registered students compared to the previous fall that had every been recorded. This drop of 3.3 percentage points represented over 3,120 students who were now passing their courses at midterm instead of failing. This has now become an effort we do every term to help our students as early as possible.
6. Take Action!
Finally, it’s important to realize that data will never be an end unto itself. Data without action is always the least desirable outcome. Every decision, action and approach you make must be towards the creation of an analytics-driven culture that consistently takes positive actions based on insights that have been informed by data. These informed actions will transform your institutional operations and help you better support your students each and every day.
You Might Also Like
