STEM: Get the Students Immersed in Data!
From Richard C. Larson
 STEM means so many things to so many people. There are stereotypes: robot design and building; using Legos to create simple structures; use of slime; computer coding. All of these are good, but a bit artificial. Things are brought to the students to bring out ideas of science, math and engineering.
Why not get young people to apply science to their everyday lives? How? Introduce them to Data Science! Have them collect data on aspects of their lives, plot the data, compute means and variances, and otherwise analyze the data both numerically and in words.
What kind of data are we talking about? Well, just about anything you can think of:
- Hours per day of watching TV, collected by time of day and day of week. Maybe also by topic or content.
- Hours per day of doing homework, collected by time of day and day of week. Maybe also by topic or content.
- Social media data, such as number of posts on Facebook, Twitter and related sites, again by time of day and day of week.
- Numbers of friends one has face-to-face conversations with, by time of day and day of week.
- The number of miles driven by one’s parents each day, by day of week. Maybe also add number of gallons of gasoline purchased and when purchased.
As one can see, this list is potentially endless and could include foods consumed, exercise, pets, and more. Some data collection can actually be automated via ‘apps’ on smart phones!
So, what does a young person do here? First, design and create a data collection sheet that will accommodate all of the data to be collected, for each application. Then figure out how to collect the data and for how long. She or he needs a sample size that will yield informative results. Perhaps two weeks to over a month would be fine, depending on the application.
Once the data collection is complete, the student learns how to use computer-based spreadsheets, typing all the data into the spreadsheet. Once that is done, the young researcher can compute means and variances (needs to learn about these) and can plot the entire distribution. If this exercise is assigned to all students in the class, then the in-class teacher can present a tutorial on spreadsheets.
These exercises could get really interesting when plotting two or more data types onto the same spreadsheet. For example, one spreadsheet might have for each day of the past three weeks data on hours per day watching TV, doing homework and posting on social media. The student can do all sorts of exploratory data analysis to see how these are related, one hypothesis being – more time with TV and/or social media, less time with homework. The in-class teacher may even want to present a tutorial on how to analyze multiple types of data together.
Finally, the proposed project will be complete when the student writes her/his findings in a multi-page interpretive report, complete with supportive charts and graphs.
Advantages of this approach: Student deals with real phenomena from her/his life; learns rudiments of data collection and analysis, a skill needed in almost all professions these days; it’s fun, and one learns a few perhaps surprising things about one’s life!
Richard Larson is the Mitsui Professor of Data, Systems, and Society at MIT. He is also the Principal Investigator of MIT BLOSSOMS.
Back to newsletter |
