Active Projects » Scratch Data Research


Scratch Data Research

One of my research interests is understanding what and how novice programmers learn as they engage in participation in online learning communities like Scratch.

In a recent study with Benjamin Mako Hill and others, we looked at learning outcomes in Scratch users as they remixed programs created by others with Scratch. While remixing has its share of proponents and detractors, our study is the first large-scale empirical analysis (examining the trajectory of over 170,000 users) of its effects on learning. We found that remixing is positively correlated with increases in the overall programming vocabularies of Scratch users, and increased levels of exposure to specific computational concepts (e.g. loops, conditionals) through remixing is associated with increased likelihood of using those concepts at a later stage.

In another quantitative study, we found that learners from non-English speaking countries tend to demonstrate use of new programming blocks at a faster rate if they are using the Scratch interface and the programming blocks in their local language (as opposed to using them in English).

Papers

Advisors & Collaborators

Mitchel Resnick, Natalie Rusk, Benjamin Mako Hill, J. Nathan Matias, William Hale, and Andrés Monroy-Hernández.