Why Data Science In Education Interests Me

17 May 2020 -

Data Science became interesting to me because I found myself spending a lot of my free time looking for patterns in data and trying to find insights that would help me make decisions for my classroom and my school. As I’ve gone through about six weeks of teaching myself how to use the tools of data science and the processes that are used in machine learning, I have had mixed feelings about how well I think they can be applied to the education space. This is my attempt to process those thoughts productively.

I’m not interested in using AI to personalize the pace and sequence of lessons for students. I am interested in using AI to help understand the factors that become barriers to learning for students. It is a common process by educators to create a list of all the things happening in a students life, in and out of school, that accumulate to become an insurmountable obstacle. It is also regularly debated when, where, and how to intervene. Still then, it is very difficult to measure the impact of those interventions. I believe that there are data science solutions that can help educators be more disciplined in this process.

I’m not interested in training a model to rank the effectiveness of teachers. I am interested in modeling out the features that can predict how long teachers will stay in the classroom and which supports help accelerate their growth. With some reservations, I think there is a way to use models like this in hiring decisions, especially in areas with drastic teacher shortages.

I’m not interested in removing grading from a teacher’s plate with AI and have them miss out on getting to know their students through reading their writing and watching them speak. I am interested in exploring ways to create journaling and reflection exercises for kids that help them process the difficult transitions in their life and use language modeling to alert teachers, parents, and mental health workers when an intervention is necessary. I am curious about ways to integrate teacher and counselor support to allow for a balance between student privacy and potential early warning systems for acute mental health issues.

I’m not interested in using models to create the most efficient ways to create new schedules to handle budget cuts and new virus related regulations. I am interested in developing simulations that allow schools to see if their new solutions are equitable for all students and sustainable for teachers.

Ultimately, I don’t think my ideas are that different from what any other educator would want to prioritize. And, maybe this is what is really going on under the surface of a lot of the data science initiatives in education. But, that’s not what I have been reading online. Either way, I hope to keep myself honest in prioritizing these things going forward in my learning and job search processes.