Design Patterns To Support Storyboards And Scenario-Based, Innovative Item Types

Abstract

Tasks that measure inquiry-based constructs, such as model-based reasoning and systems thinking, often require students to complete multiple, cognitively demanding steps, frequently in the context of a larger scenario meant to reflect a “real-world” situation. Designing these tasks for use in large-scale assessment requires task writers to integrate multiple content and inquiry targets across a series of scenes, each containing multiple embedded items, and to do so in a way that the resulting scores contain little measurement error and support valid inferences about student performance. Design patterns are research-based, representational forms that can help assessment developers organize and integrate information about tasks in terms the focal skills to be measured, the behaviors that reveal the focal skills, and the task features that will elicit the desired behaviors. In this paper we report on current research in which we are examining the extension and application of inquiry-based design patterns to guide the development of complex science tasks for use in a statewide science assessment. As an example, we describe the use of a design pattern representing the practice of model revision to develop a task asking students to consider the introduction of an invasive species into a new ecosystem and to revise a food web to accommodate for the introduction of the new species.