Developing a kindergarten computational thinking assessment using evidence-centered design: the case of algorithmic thinking

• Published in: Computer science education Vol. 31; no. 2; pp. 117 - 140
• Main Authors: Clarke-Midura, Jody, Silvis, Deborah, Shumway, Jessica F., Lee, Victor R., Kozlowski, Joseph S.
• Format: Journal Article
• Language: English
• Published: Norwood Routledge 03.04.2021; Taylor & Francis Ltd
• Subjects: algorithmic thinking; Algorithms; assessment; Cognition & reasoning; Component reliability; Computation; Computational thinking; Design analysis; early childhood; Early childhood education; Educational evaluation; evidence-centered design; Kindergarten; Measures (Individuals); Reasoning; Spatial Ability; Test Reliability; Test Validity; Thinking Skills; Young Children
• ISSN: 0899-3408; 1744-5175
• DOI: 10.1080/08993408.2021.1877988

Background and Context: There is a need for early childhood assessments of computational thinking (CT). However, there is not consensus on a guiding framework, definition, or set of proxies in which to measure CT. We are addressing this problem by using Evidence Centered Design (ECD) to develop an assessment of kindergarten-aged children's CT. Objective: To present a design case on the development of the assessment, specifically the algorithmic thinking (AT) tasks and to share validity evidence that emerged. Method: We focus on the AT sub-component of CT and present the principled assessment design process using ECD. Findings: Our operationalization of CT includes spatial reasoning as a sub-component. Pilot results showed an acceptable internal consistency reliability for the AT items and critical design decisions that contributed to validity evidence. Implications: An important contribution of this work is the inclusion of spatial reasoning in our definition of early childhood CT.