How would you like to get more from your data collection efforts without doing more data collection?
I’m talking about getting more out of your surveys, rubrics, engagement data, and even your event attendance tracking information without asking more questions or taking up more of students’ time. And how would you like to do it without having to maintain or rely on your own records of data and information about students?
Let’s talk about data integration. Most institutions have multiple data sets stored in various places and within various systems: Student enrollment and records information (including demographics, address, transcripts, financial aid, and more) is stored within your Student Information System (SIS), class participation and grade data is stored within your Learning Management System (LMS), student participation or engagement data is stored within a student engagement platform, and your assessment data may be stored in a dedicated Assessment Management System (AMS).
Through a little technological magic, we can link and connect these data sets. And by “we”, I mean that you’re likely going to need to bring in your fellow campus data guardians. While some collaboration may be needed, connecting these data sets is possible and beneficial for multiple parties.
I’ll talk more about the how later on; first let’s talk more about why you would want to do this.
Top Benefits of Connecting
Let’s focus just on student affairs professionals connecting to their Student Information System (SIS). If you are able to tap into this, you could shorten your data collection efforts on forms and surveys and eliminate the need for students to self-report information about themselves, such as their majors, identity demographics, and class years.
All you would need is to collect a common identifier (such as institution email or ID number), which will enable you to pull in that student’s information from the SIS.
Beyond reducing the number of questions you’ll repeatedly ask students to answer, integrated data systems can help ensure more accurate and consistently collected data.
Consider these scenarios:
- The Office of Student Activities needs to know the class year of students participating in events in order to better gauge the campus’s programming needs.However, different professionals within that office have been collecting that data in different ways for the programs they’re in charge of. One person collects class years via first-year through senior options, while their coworker surveys students on their credit ranges, and a third coworker asks if attendees are undergraduate or graduate students.The office is left to reconcile these various response options, which may or may not coincide with institutional classifications.
- For Women’s History Month, Career Services wants to know the identities of students attending events in order to gauge the effectiveness of the office’s marketing campaigns.
In one instance, they ask attendees to respond to question asking about sex, and, in another, they ask about gender. As those are different data points, they can’t be reconciled. Moreover, if the institution only has one of the two on file for students, there is no accurate way for Career Services to know what percent of the total given student population they are attracting.
I’m sure you can identify with these data dilemmas and pain points. The confusion around institutional definitions or classifications, as well as accuracy and consistency of student information, can all be taken care of by tapping into the SIS instead of asking for students to repeatedly self-report.
Plus, as I alluded to in the career services example, when we have access to institutional data sets and utilize them, we gain insight into equity considerations. We can accurately know our student populations, making it easier for us to disaggregate (separate complete data set into distinct groups of data) our collected data related to engagement, satisfaction, and performance.
With disaggregated data, we can see where the identities of students significantly differ from our goals or learning targets. When disaggregating data or looking for differences related to identity, it is important to either compare identity populations against an institutional target or set population-specific targets. Be careful not to compare identities against one another, as that can make it seem like you view one population as superior over another.
All of those examples are just pointing to connections within the SIS, but if you add in the LMS, AMS, and other data systems, you’ll have a recipe for some awesome collaborations. Imagine being able to talk about student population behaviors or trends that you can link to classroom engagement/performance, identities and demographics, major/credits earned, and more. You could invite student engagement and advising staff, the registrar, and faculty to consider things like course load, completion, and progress of involved students.
Such rich and robust data connections might assist you in making progress toward understanding big, complex concepts — like the connections between a student’s sense of belonging and academic success.
Steps Toward Intergration
Now that you have an idea of application and use cases for integrated data systems, here are a few steps and considerations to make this a reality at your institution.
- Assemble a squad. Bring together the following areas to make this happen, if each indeed exists at your institution: IT staff, registrar, financial aid, institutional research, and business intelligence.
- Create space for collaboration. Once the data is connected, you should keep engaging and/or bringing in more people or departments to add to a conversation about strategizing and supporting the use or applicability for these data.
- Make time for review. Carve our time on your calendar to dedicate toward reviewing the data available to you. It is true you may get some time back from not having to bring together disparate data responses that were self-reported. However, you’ll be afforded much more data than you were likely collecting in the first place, so you need to allocate time to review those data in relation to questions you need to be answered.
- Maintain a focus. Having access to more data can be exciting and offer all sorts of new insights. It is important to remember your priorities so that you are not fishing or allowing your curiosity to set you swimming in deep data waters. Reviewing large data sets without a plan cant sap you of time and pose questions that are unrelated to your immediate purpose and needs.
Beyond what is presented here, you may have to go through more steps to get the data connected. Likewise, there may be some data sets that are easier or more likely to be connected than others.
Any data connection is a good one, so take what you can get. And don’t be surprised if, when you go asking to connect, you find out that many data sets are already connected! In that case, explore how you can leverage these connections for your engagement or assessment purposes…and then tell your coworkers to spread the word.
As my institution, National Louis University, has worked out the kinks in connecting the data and scheduling how often it refreshes, we’ve benefitted from being able to connect student experience data with grades (from our LMS) and student identity information (from our SIS). This has given us a robust picture of the student experience. We’re able to assess their behaviors and performance inside and outside of the classroom, all while being able to disaggregate or focus on particular demographics for insights or trends among populations.
It has been particularly helpful at our institution to check in monthly as a Business Intelligence group. This group consists of our BI person, the director of IR, the coordinator of our Banner system (SIS), and the many data-oriented folks like me who have a stake (and access) with these connected data sets. This committee-like setting allows for sharing with one another if there is new data available, discussing data initiatives (like putting together a data dictionary), or brainstorming solutions regarding difficulties we have experienced with certain data pulls, analyses, or visualizations.
In bringing people together to discuss these data sets, we have found that we all exist on a spectrum of knowledge and familiarity with the systems and skills necessary to work with these data sets. As such, we developed user groups to build capacity around tools like Power BI for analysis and visualizations, as well as skills like using SQL for data queries. We save and share our training and professional development materials so that when new people join the groups or gain access to the data, they have content to reference and learn on their own.
Enough about me; we’d love to hear from you! For folks who may have done this already, do you have any additional tips, tricks, or lessons learned? Let us know by connecting with us on Twitter @HelloPresence and @JoeBooksLevy.