Last week, a colleague of mine looked at me – with eyebrows raised in panic – and said, “But, I’m not a math person!” I suggested she do some data analysis to find out if a new strategy she tried was working. That sounded like complex mathematical analysis to her, though, and she panicked.
In reality, humans are hardwired to collect, interpret, analyze, and synthesize data. We start as babies when we mimic adult facial expressions and sounds. Every middle schooler who’s tried to “fit in” uses data to understand what’s cool and what’s loathed by the Queen Bee of the 8th grade. For many of us, this data collection is instinctive and automatic, informed by years of observation and analysis of the world around us. But because we’re not looking at numbers and spreadsheets, we don’t typically think of it as data-informed decision making.
At PelotonU, we harness both “instinctive” and more traditional data to understand how our students are doing, and therefore, how we’re doing. Our mission is to help working adults earn a college degree on time with little-to-no debt. Alongside deeply relational in-person coaching, we leverage ultra flexible, high-quality online degree programs in which a student sets their own schedule and pace. Each of our college completion coaches uses what I’m calling “instinctive data” (Hmm… that student usually texts back right away, but I haven’t heard from her in a week. Seems like something might be wrong – I should check in again.) and quantitative data (How many days since that student last submitted an assignment? What’s her projected time to graduation?) to monitor student progress and well-being.
Our coaching team examines student data on a weekly basis, both as individual coaches and as a group. One tangible practice we’ve embraced is checking our confirmation bias. Often, a student is thriving early in the program so we file them under “successful” in our brains, making it easy for us only to notice data that confirms what we already believe: this student is doing well. Through weekly data analysis, however, a coach finds that the student hasn’t submitted an assignment in nearly a month, and they’ve added 6 months to their projected graduation date: in other words, the student is not thriving anymore.
While data-informed decision making has been core to PelotonU since our inception, we’ve also evolved our practices with time. As we see concrete evidence of what was working about our program or what wasn’t working, we were compelled to make improvements to serve current students better or to being serving students we’d unintentionally left out. With that, we start to ask new questions of our program, collect new data, and learn new things about how we’re doing by our students.
Whether you’re brand new to data, building buy-in among colleagues, or leveling up in sophistication, we’ve found time and again that clarity is king. We must be clear on what question we’re asking and why – and to this, the answer must always be “to make sure we’re doing right by our students.” Then we must be clear about what data we need to answer our question, who collects it, how we’ll use it, and how we’ll maintain it. When done well, data-driven decision making both hones our instincts and shows us tangibly that we are caring well for those we serve.