Location: Ballroom C+D+E
This is a roundtable in the Cyberlearning 2017 Roundtable session.

Integrating computer science in science instruction: how do we know its working?

Gillian Puttick

Our design-based research project is working with 8th grade teachers to integrate computer programming in Scratch in their science instruction. With respect to game design, science, and computing, constructs we are measuring include student attitudes, motivation, and self-efficacy, as well as science content knowledge and computational thinking (CT). Y2 efforts are directed towards better integrating systems thinking in the curriculum, and towards expanding our measures for CT beyond those that are apparent in student programming. The boundaries between CT and systems thinking are fuzzy, and research findings from interventions to develop CT in non-computer science disciplines more generally are sparse. In this roundtable, we will explore what are some models of CT, what the major components are, and how researchers in various disciplines are measuring it.
Project: Homepage, NSF Award #1542954 – Building systems from Scratch. Research on the Development of Computational and Systems Thinking in Middle School Students through Explorations of Complex Earth Systems

Models of Learning Progress in Solving Complex Problems: Expertise Development in Teaching and Learning

Min Kyu Kim
This study proposes that learning is a process of transitioning from one stage to another stage within a knowledge base that features concepts and relations. Drawing on the theories of expertise, this study explored two different models of learning progress (i.e., three- and two-stage models) in the context of classroom learning and identified a model that was a good fit to the data. Participants in this investigation included 136 students and 7 professors from six different universities in the United States. In order to detect and validate stages of learning progress in participants’ written responses to an ill-structured and complex problem scenario, this study utilized Exploratory Factor Analysis (EFA) and the Continuous Log-Linear Cognitive Diagnostic Model (C-LCDM) method (Bozard, 2010). The results demonstrate that the three latent classes matched the three stages of the three-stage model. This study provides an account of a diagnostic model of learning progress and associated assessment methods, yet further studies are required to investigate different conditions.
Project: Homepage, NSF Award #1623561 – Inclusive Design for Engaging All Learners

Hint avoidance in programming

Tiffany Barnes

We have been researching the use of automatically generated hints for open-ended novice programming tasks. However, many students do not use the hint system. Current results suggest that the quality of the hints directly relates to whether or not students follow the hints or request more, but many students never request hints at all. We will discuss this and ways we are working to encourage more hint usage.
Project: Homepage, NSF Award #1623470 – EXP: Data-Driven Support for Novice Programmers