New Research Report Calls for Better Assessments for Computer Science Education

Posted by Chris Stephenson, Head of Computer Science Education Programs/span>

Editor's note: Ensuring the appropriateness, value, and impact of our efforts in the computer science education space first requires an understanding of the issues which broadly impact the discipline, its practitioners and its students. This article is part of our ongoing effort to explore those issues and share our learnings along the way.

A new research report published by the Computer Science Teachers Association (CSTA) — and funded by Google — highlights the need for valid and reliable sources of assessment to measure student learning in computer science education.

Sowing the Seeds of Assessment Literacy in Secondary Computer Science Education details the results of a landscape study aimed at determining the challenges U.S. high school teachers face when examining student understanding of computing concepts and identifying current models for computer science (CS) assessment. The study was conducted over twelve months by the CSTA Assessment Task Force chaired by Aman Yadav, Associate Professor of Educational Psychology and Educational Technology at Michigan State University. It involved in-depth interviews with U.S. computer science practitioners with a wide range of teaching experience.

The study concluded that while computer science teachers use a variety of formative and summative assessment techniques and rely on an assortment of sources (test banks, colleagues, even their own undergraduate CS courses), they face a number of challenges finding valid and reliable assessments to use in their classrooms. It also highlighted the potential for variability in how students approach and develop algorithms as one factor that makes assessment especially challenging and time-consuming. 

The report calls upon the computer science education community to assist in the development of more and better assessment tools and strategies and suggests the following next steps: 

• Develop valid and reliable assessments aligned to the CSTA K–12 Computer Science Standards. 
• Develop valid and reliable formative and summative assessments for programming languages beyond Java, such as Python, C#, etc. 
• Develop an online repository of assessment items for K–12 computer science teachers.
• Develop a community of practice surrounding the use of assessment in computer science classrooms. 
• Design and deliver professional development to increase K–12 computer science teachers’ assessment literacy. 

Yadav says he hopes that this study will focus the computer science education community’s attention on the importance of valid assessment of student learning and the pressing need for better and more computer science assessment tools and strategies.