The Design and Effects of Data Science Workshops for Educational Researchers

Abstract

Though there is growing interest in using novel methods to analyze novel sources of educational data, numerous barriers and challenges exist for researchers and analysts seeking to learn to use educational data science methods in their work. We report on our design, development, implementation, and research of two data science workshops focused on using R and RStudio for 44 early career educational researchers with limited and varied EDS backgrounds. Through a qualitative and quantitative analysis of pre- and post-workshop surveys, we found that participants in educational data science workshops are concerned and even fearful, but also possessing of resources in the form of prior degree programs and work and teaching experiences that relate to analyzing data. Participants reported that pedagogical factors—coding in R together and talking through the code and its output, being patient amidst errors, and explaining technical concepts in an accessible and rigorous manner—were strengths, while the need for more time and exposure to data science were possible improvements pointed out. An implication of this work is that providing greater opportunities for researchers at all career stages interested in developing knowledge and skill related to educational data science has the potential to not only broaden access to potentially useful capabilities but also to shape and define educational data science in a broader—and ultimately better—manner.