To Teach Quantified Self, First Know Thyself

Michael: Our school is Summit Shasta over in Daily City right next to San Francisco, and fun fact, even although both of look like high schoolers only one of us actually is. So, quick trivia is for you to figure out which one of us is still in school.

Alex: It’s him, it’s him.

Michael: All right, so Alex has been great on this presentation. He just lost some points. Anyway, I am a math teacher.

Alex: And I’m a fourth-grade student.

Michael: So here’s a story. Our school is all about experiential learning through projects. For example, in US government they emulate their own mock congress.

Alex: In chemistry, we created bio-fuel.

Michael: But in math class, there is this intense pressure to prepare seniors for either a very rigorous AP Calc, AP Stats exam. So last year we decided to forgo projects and solely focus on content.

Alex: Basically, last year Michael’s class was really boring, super hard, and 80% of the students failed the exam:

Michael: So, we really got the worst of both worlds. Students weren’t super excited about the class, didn’t learn a lot and it really made no sense for us to make that move considering stats is all about real world data analysis.

The good thing that our school really embraces freedom and creativity for both teachers and students.

Alex: For example, students can create electoral courses in our study program, and last year, I helped my AP chemistry teacher create resources and structure of the course.

Michael: And when I heard about Alex’s awesome work in chemistry, I thought hmm, free help. So, I asked him to team up with me for stats this year. And the result was, basically we would take the curriculum and flip them around instead of having seniors learn a bunch of mathematical equations and formulas to memorize and then dummy stats to work on. We would first have them collect personal meaningful Quantified data, and then they would be excited to learn about the math.

Alex: Our first project is about Quantified Seniors. Basically, we would do wide analysis, pick three habits and track that over a month and make some predictions and analyze those behaviors through univariate and multivariate methods. Also, this perfectly ties in with the fact that I’m a senior and much of my peers are as well, which means that it’s a time of self-reflection and deep thinking about who we are.

Michael: As they apply to schools. So, we decided to run a pilot of our own curriculum over the summer to test drive what students might be interested in testing. So we ran things like productivity, nutrition, sleep, fitness etc.

Alex: You know how weird and shocking it is to hear your own voice for the first time, that’s what this pilot felt like.

Michael: So I just want to call attention, I just want to commend Alex for his nearly a whole day tracked on YouTube in one week. That’s like a YouTube Sabbath.

Alex: I’m not embarrassed to show you my awesome productivity percentage. Here at Shasta it’s all about growth and improvement. I mean, look at that 7.4 increase from last week. I’m guilty though. I fell victim to the Fortnite gaming movement, and over the course of 36 days I did some variant analysis and found that the medium was roughly 1.5 hours a day. It being a right-hand distribution also met that not all my time was wasted on the Fortnite bandwagon. And I think the important thing to note is that there are also some outliers, like three all-nighters.

Michael: Right, these all-nighters weren’t technically outliers because they because they fell within the 1.5 of IQR test, so he just lost some more points.

Alex: cancel part of the project for seniors was not just understanding transcendental one variable, but a relationship between variables and how they affect each other.

Michael: So we thought we could hear some variables, one being my productivity working on curriculum as tracker RescueTime, and use other variables to explain it. So, my nutrition intake via carbs or sugar or my hours spent working out or sleeping.

And so the one model that was pretty interesting was sugar versus productivity. The Y intercept over here, places me at a baseline of roughly 3 hours of productivity a day if I eat no sugar. Then for every gram I consume there after I lose about half a minute of productivity.

Now, this narrative is nice and pretty when it comes to our assumptions of what we think sugar does to people, except when we ran through the fit of the model, we get something that’s not so pretty. So basically in r-square of .016 means that about half a percent of my change in productivity can really be explained by my change in sugar intake.

Alex: I thought hey, why don’t we put Michael’s variables into one big multivariate model to help better explain things.

Michael: And we were wrong, it didn’t help. It was worse. You can see here that sugar is close to increase my productivity and sleeping more and working out more actually decreases it and then this P value I won’t even talk about that. Basically, at this point we were like. So we were thinking when the seniors get to this point in the project and got two weeks we’re just going to wing it. We’re going to tell them, chalk it up to experimental design flaws or the stochasticity of human nature. So here’s…

Alex: Fast forward one school year and we started data tracking for the class. Here’s some of the interesting things that they’ve been tracking. They’ve been tracking the number of Snap Chats they’ve been sending in a day, the duration of call length to extended families and how much money they are actually wasting on food.

Michael: We’ve also got Netflix binges per week, and some pretty cool stuff like word count of a gratitude journal or a gratitude letter sent to friends. And on the flip side there is roasts, which is a teenage code for insults that they send to their peers.

So, here’s some titbits that we learned from the project. Basically, in terms of our own personal habits, vices, and sugar is one of mine follows a bimodal distribution. So round a whole bunch of days I eat super good, I eat no sugar at all and then a couple of days or a whole bunch of days where I’m just like on the floor staring at for empty parts of Ben and Jerry’s.

Alex: As from my Internet habits, I’m starting to realize that maybe Google’s a scam. Maybe they are not looking out for my best interest, but rather their recommended content is kind of BS.

Michael: And for general life principles, Peter Drucker had this quote that goes, what gets measured gets managed. So full even though most of our aggression models were pretty inconclusive. Just simply tracking our data, we found our students and ourselves had become much more intentional about our behavior.

Alex: I think a bunch of us just think that math just sucks and it’s boring, but all we want is something that’s meaningful and useful to us.

Michael: And finally, I think it would have been impossible for us to keep up with all this data tracking without the support of each other building this curriculum. So, I think, we’re trying to figure out how can we better work social pressure and accountability into the senior project, and like maybe leverage their grades off of that.

So, that is our work in a nutshell, and I think one of the assumptions that I’m aiming for this project was that, teenagers either don’t want to change or didn’t feel they had the power to change their own lives. But I was wrong.

Alex: We’ve started to realize that we can not only shape ourselves but also the environment around us. And part of our campaign is understanding this for the classroom. And scientific research has shown that there is a 3 to 5% increase in test scores, increased calorie burn, and actually, up to 15% increased engagement per hour.

Michael: So if you are compelled to either share or support our campaign, the link is right there tinyurl.com/qsdesks and we thank you so much for your time.

Alex: And money.