A Journey into Data Science

Josh Johnson
3 min readSep 7, 2020

From Educator to Data Scientist.

Photo by NeONBRAND on Unsplash

My middle school students frantically drew blurbs of text on the margins of large sheets of butcher paper. In groups of four and each with their own colored marker, they discussed, deciphered, and paraphrased articles of the U.N. Declaration of Human Rights to decode their meanings. The would use this understanding as part of their projects on modern-day slavery and human trafficking.

I took comprehensive observational notes and counted the contributions of each member by their marker color. At the end of the class period, the students took a survey asking how they enjoyed the group work, whether they felt they and their peers contributed their fair share, and answered an assessment question.

Student responses to structured group work
Image by Author

We can see here that student enjoyment of their group work experience is highly correlated to the perception that everyone contributed, as is a feeling of inclusion.

I took data on four different group work sessions in two different cohorts and spent a weekend crunching the numbers in Excel. Counter to my preliminary research, it seemed that students preferred less teacher imposed structure in their groups. However, my data supported the theory that increased enjoyment correlated with improved learning outcomes (as evidenced by the assessment question).

Image by Author

This chart shows that while students did not report enjoying structured group work (Jigsaw structure) more, learning outcomes were improved.

I lost myself in graphs and charts, enthralled by insights flowing out of them about my students, about learning, and teamwork. However, I was painfully aware of the tiny amount of data I had to work with. I had only 70 observations with four features over 4 periods.

My detailed findings can be found here.

I wanted more. I wanted more data, I wanted better tools, and I wanted to know, not rely on thin and outdated theories and scant or barely relevant empirical evidence. Most research on learning is done with college students, not middle schoolers. How could I discover the best ways to support group work?

Image by Author

Students reported on their experience and perception of the group work experience.

My passion turned me toward data science. After 15 years of teaching, I was ready to leave the classroom and dive into data. I started with Coursera classes, and once again lost myself in linear regressions and recurrent neural networks.

I’m ready now to do real work, get real data, and make a difference. I’m ready to join a data team and answer big questions. I want to improve learning for millions of children, not just the thousands I might impact in the classroom.

After some research, I decided to enroll in the Flatiron Data Science Intensive boot camp. Now I’m learning with experienced instructors, devoted classmates, and widely networked career coaches to find the right place for me to change the world.

What’s your inspiration?

--

--

Josh Johnson

I'm a data scientist with a background in education. I empower learners to become the folks they want to be.