Before I get into this post, I want to make sure that everyone gets the tongue-in-cheek nature of the post title and image, so please check the meme reference if you’re not sure. Now that I’ve got my implicit disclaimer out of the way, let’s start by talking about sports analytics. I’m a bit of a fanboy when it comes to sports analytics. It’s a mashup of two of my true passions. I’m the data nerd who plays fantasy baseball and fantasy football every year. I haven’t been to the Sloan Sports Analytics Conference, but I’m sure I’ll make it some day. I’m a runner and I chart out my race times year over year. I have a wi-fi enabled scale that tracks my weight. If the Arizona Diamondbacks’ Director of Analytics Dr. Ed Lewis happened to come knocking at Blue Canary’s door, I wouldn’t turn him away.
Analytics in sports (especially in baseball) has been on a steady rise, with much of that credited to the Moneyball influence. It has even gone so far that this past year, Fox Sports aired an analytics-driven broadcast of an entire MLB playoff game. With that rise, though, has come the inevitable counter-attack criticizing analytics.
A CBS Chicago reporter penned a piece last year in response to a high profile criticism of sports analytics. The critical article lamented the spread of sports analytics into places that the author didn’t think it belonged. His best quote from the article was:
Must all the intangibles be sucked from our games until all that is left is spreadsheets and blinking computer screens?
The rebuttal article replied by taking the stance that the goal of analytics is not to replace the “human factor” of sports with computers. Instead, analytics can complement the humanity, emotion, and intangibility of sports in a way that would make even the most old school fan enjoy the sport a tad more. The author says:
Analytics are an attempt to better understand the games and add to them, not subtract. I still don’t understand much of the science very well, but I know it’s a well-intentioned movement that aims to enhance and not hinder…
Here’s where I see a wonderful connection to analytics in education. As a practitioner and one who tries to keep a close ear to the #edtech #analytics Twittersphere, I see a fair amount of pushback against analytics. Some of it is justified — I’ll be the first to own up to the fact that there are players in the space whose misuse of terms like “big data”, “magic”, and “secret sauce” don’t help to illuminate analytics in a good way. However, there are a good number of practitioners who follow the same path as described in the sports analytics article. They are passionate professionals who are well-informed and well-intentioned. Their goal is to empower institutions, faculty, and students by leveraging data to help drive student success. The goal should never be to take the human out of the equation. Bad things happen when you remove humans from the loop. Sometimes, really bad things happen.
To be fair, not all of the criticisms of analytics are unfounded. Continuing on with the sports analytics connection, there’s a wonderful article about a basketball coach’s take on the validity of data in analytics. Stan Van Gundy, coach of the Detroit Pistons, was referenced in the piece:
Van Gundy really does like the additional available data — he just needs to be able to trust that whoever is compiling it has the same standards basketball-wise that he does.
Van Gundy goes on to give an example using the basketball “pick and roll” maneuver. He says that the pick and roll is used to accomplish different things (scoring, positioning, setting up the next play). If all the statistics do is to count the number of pick and rolls, then the analysis lacks depth and validity.
That’s context in analytics…especially in academic analytics. You can’t treat all assignments the same. You can’t treat all discussion forum posts the same. I once did a predictive model where we used missed attendance in a given week as a proxy for attrition. That was a valid proxy measure for the online classes, but it turns out it was meaningless for face-to-face classes. Students in face-to-face classes would sometimes take their one allowed absence as a “mulligan”. There was no intent of dropping the class, they were just working adults who would gladly take a penalty-free day off of class (as opposed to online students, where not participating in a seven-day window was usually a precursor to dropping the class).
So for those who tend towards the critical side of analytics, I hope that this comparison to sports analytics has helped to shine a positive light on the world of data. Analytics are not a cure all, but they are not a curse either. Analytics won’t fix poor <teaching/instructional design/student motivation/etc.>, but it can move things from good to better.