Easing learners into data science via visualization of concepts and computations

Abstract

In my research, I will explore the potential of adapting the literature on algorithm visualization and visual analogy to the teaching of concepts and computations in introductory data science. Using computerized tutorials informed by the extensive literature on multimedia principles already available, I will explore if a visual pseudo-code facilitates simulation and application of data algorithms, beyond the facilitation afforded by mathematical notation. The research will combine between-subject classroom interventions with distance learning and within-subject laboratory studies. The dissertation will give fine-grained evidence on which types are more effective and how it could be implemented in the community at large. I am primarily interested in discussing methodological issues relating to how to make comparisons in such a multi-dimensional design space.