An Approach for Assessing Computational Thinking in Mathematics and Science Classrooms

• In Event: 61.039 - Assessing Computational Thinking: A "Landscape" Symposium About Where We Are at and Where to Go

Abstract

Objectives
As computational thinking and its associated skills gain prominence across K-12 educational settings, there is growing need to be able to assess learners’ emerging understanding of key concepts and essential practices (Tissenbaum et al., 2018).

Background
Historically, computational thinking instruction, if taught explicitly at all, has been embedded within classrooms and is tightly coupled with concepts and skills associated with programming. As such, many computational thinking assessments build on this tradition and rely on conventional programming questions and concepts as part of the assessment (e.g. Brennan & Resnick, 2012; Grover, Pea, & Cooper, 2015; Koh, Basawapatna, Nickerson, & Repenning, 2014; Werner, Denner, Campe, & Kawamoto, 2012). While this approach makes sense in many contexts, the growing diversity of domains and classes in which computational thinking is being taught highlights the limitations of such an approach. For example, the Next Generation Science Standards (NGSS) including computational thinking as one of eight core scientific practices (NGSS Lead States, 2013). While things like algorithmic thinking or problem decomposition fit well into science classrooms, conventional computer programming is rarely taught as part of the science curriculum. As such, there needs to be a way to assess computational thinking that is decoupled from programming.

Methods
Our conceptualization of computational thinking is based on Author 05A et al.’s (2016) Computational Thinking in Mathematics and Science Practices Taxonomy. This framework breaks computational thinking down into four overarching around contemporary science and mathematics: 1) Data Practices; 2) Modeling and Simulation Practices; 3) Computational Problem Solving Practices; and 4) Systems Thinking Practices. Each of these categories is composed of specific practices, which collectively define computational thinking as we understand its application in authentic mathematics and scientific contexts.

Results & Next Steps
To evaluate students’ emerging computational thinking abilities in math and science classrooms we developed a series of interactive, online assessments (Author 05 et al., 2014). In designing these assessments, we developed two design principles. First, each assessment needed to be situated in a mathematical or scientific scenario so as to be authentic with respect to the nature of the computational thinking practices students were being evaluated on. At the same time, we were careful to make sure the assessments were not dependent on specific mathematics or content knowledge (so that they could be used across a wide range of curricula). Second, the assessments needed to have learners enact the computational thinking practices they are being assessed on. The resulting assessments embed computational tools, such as interactive visualizations, simplified programming tools, or models of scientific phenomena, alongside questions that we ask learners to use in order to answer the posed question.
Collectively, our goal with this line of research is to provide an approach to assessing computational thinking as it resides in learning contexts beyond computer science. In doing so, we seek to help educators and researchers better characterize students’ emerging understanding of and competencies in computational thinking practices in mathematics and science contexts.

Author

• David Weintrop, University of Maryland - College Park