Concept Inventories: Design, Application, Uses, Limitations, and Next Steps

Abstract

The idea of a concept inventory or concept test, as opposed to a conventional test, emerged during a period when education researchers and educators began to focus on identifying students’ prior conceptions that can act as barriers to learning science. The typical concept test is multiple choice in format and focused on widely applicable concepts rather than facts. The incorrect choices, known as distractors, represent scientifically problematic ideas that students actually hold, presented in students’ own (recognizable) language. Often multiple questions are used to target common or linked concepts. The validity of a concept test arises primarily from its distractors. In fact, distractors that do not represent or resonate with what students’ actually think undermine the meaning of student scores. While improvements on concept test scores are often used as a proxy for overall “learning gains” within a course or a curriculum, this is problematic, particularly when the same test is used in a pre-/post-model. A well-designed concept test can reveal the presence and/or persistence of conceptual confusions, but more accurate measures of learning are necessary to reveal what students can do with their knowledge. That said, evidence of conceptual confusion at the start (or the end) of a course can be critical when one considers the content presented and the types of formative learning activities that might help students develop an accurate working understanding of core disciplinary concepts.