Constructive (or Destructive) Ambiguity: The Benefits and Drawbacks of Definitional Imprecision in Computational Thinking

• In Event: Dynamics of Computational Thinking: Theoretical and Practical Debates in Teacher Training

Abstract

Motivation
Scholarly interest and support for the integration of computational thinking (CT) in elementary school curricula has substantially grown since Jeanette Wing (2006, 2008, 2011) argued that CT should be integrated into every child’s “repertoire of thinking skills” from the earliest years of schooling. Increasingly, scholars have argued that teaching CT in elementary schools can improve students’ learning in diverse subject areas, foster creativity and, in the long term, increase and diversify enrollment in STEM careers (Barr & Stephenson, 2011; Grover & Pea, 2013; Voogt, J., et al., 2015; Yadav, Hong, & Stephenson, 2016). Despite extensive support, the field has not agreed upon a clear definition for CT, particularly in elementary education. Scholars and practitioners lack a precise and shared understanding of what skills constitute CT (e.g., abstraction, algorithms, decomposition), and further ambiguity characterizes the definitions of what the practice of teaching and learning CT should entail for teachers and students (Angeli et al., 2016; Shute, Sun, & Asbell-clarke, 2017). In this context, we aim to elaborate on the consequences of this definitional imprecision for CT teacher-training initiatives in three higher education institutions. In particular we ask:

1. How and why do definitions of CT vary across and within institutions?
2. What are the implications of this variation for design, implementation, and participant buy-in?

Data and Methods
Data were gathered as part of an ongoing fully integrated mixed-methods study of CT teacher-training initiatives in the Northeast of the United States. Our qualitative data are primarily derived from over 50 semi-structured interviews with computational thinking and computer science experts, representatives from academic institutions, and pre- and in-service teachers participating in CT training. We supplement these data with classroom observations and document analysis of the resources used by the participating institutions, progress reports, and instructional materials. These qualitative data were then triangulated with surveys of all pre- and in-service teachers participating in the three initiatives (n=103, 74.6% response rate).

Findings and Implications
Each of the institutional design teams began with somewhat similar definitions of CT across institutions, emphasizing that CT was a problem-solving process that included decomposition, abstraction, and pattern recognition. However, in interviews we uncovered substantial variation in how CT definitions evolved, shifted, and were used across and within institutions during implementation. These shifts and inconsistencies supported buy-in, and one organization even shifted their materials to feature a less precise understanding of CT with the explicit purpose of increasing interest in their program. However, this definitional ambiguity also led to confusion when participants, designers, and implementers expressed conflicting beliefs about the difference between CT, mathematical thinking, and problem-solving. Many respondents conceptualized CT as “just problem solving,” provoking concerns that definitions of CT were oversimplified and missing key components. This also made it difficult for teachers and teacher candidates to identify when they--and their students--were resolving a problem using CT practices, and when they were not. Thus, we find that definitional imprecision seems to be productive in terms of capturing interest and securing participant buy-in but problematic for constructing coherence in implementation.

Authors

• Melissa A Lyon, Brown University

• Carolina Snaider, Teachers College, Columbia University

• Douglas Ready, Teachers College, Columbia University