Three research-based approaches to the design of physics activities
A variety of activities are commonly used in college physics courses including lab, tutorials, and studio curricula. Instructors must choose among using research-based activities, designing their own activities or modifying existing activities. Instructors' choices depend on their own goals and the goals of activities from which they are choosing. To assist them in developing or modifying activities for their situation, we examine research-based activities to determine their goals and the features of the activities associated with these goals. Since most activities ask students to perform tasks to assist them in learning, sixty-six activities from eleven different research-based curricula were coded for student actions. The coding scheme containing forty-nine codes in ten categories was developed from a subset of activities, interviews with some of the activity designers, and recommendations from the American Association of Physics Teachers 2014 lab report. The results were examined using k-means cluster analysis revealing three design clusters. We label these clusters Thinking like a Scientists, Learning Concepts, Building Models. These three clusters reflect diverse design goals. In the Thinking like a Scientist cluster, activities emphasize design of experiments by students, discussion, error analysis, reasonableness checking, supporting claims, and making assumptions or simplifications. The Learning Concepts cluster focuses on prediction of results and experimental observations. The Building Models cluster emphasizes discussion and answering physics or math questions that do not use collected data. This work connects common features appearing in physics activities with the goals and strategies of the designers. In this way it may provide instructors with a more straightforward way to creat activities which achieve their desired outcomes.