ABSTRACT
This chapter describes the use of ECD as a systematic method to guide assessment development so that a coherent validity argument underlies the inferences made on the basis of student performance. It sets forth, for both assessment experts and science educators, examples of the logic and the validity arguments needed to support the design and development of life science measures that document the knowledge, skills, and practices that students have attained at the end of their educational experiences. The number, nature, and granularity of the proficiencies and the way they interrelate are determined by the purpose of the assessment—a single student model variable to characterize students’ overall proficiency or a multidimensional model of conducting inquiry in the life sciences, as part of an annual statewide test. As assessments of life science evolve toward computer-based delivery, automated scoring will become an increasingly critical part of the development process.
