Creating Curriculum-Embedded, Performance-Based Assessments for Measuring 21st-Century Skills in K-5 Students

• In Event: Performance-Based Assessment: Something Old, Something New

Abstract

This paper will share the author’s experiences working with a large and diverse school district to design curriculum-embedded, performance-based assessments (PBAs) that measure 21st century skills in K-5 students. These assessments are designed to integrate seamlessly with classroom instruction; measure skills such as critical thinking, creativity, and collaboration in a developmentally-appropriate way; and provide information useful for supporting classroom teaching and learning. The paper will begin by outlining general design principles that support the construction of these PBAs. For example, assessments are intended to be easy to implement in the classroom; thus, tasks include a variety of supporting materials needed to administer and score student performance. PBAs are distributed across multiple subject areas (e.g., reading, writing, math, science, social studies), given evidence that many 21st century skills are domain-specific. In order to create tasks that fit seamlessly with instruction, PBAs are developed out of existing instructional modules. Finally, PBA tasks include general (rather than task-specific), rubrics. Teachers will conduct their own scoring, and it is believed that teachers will more quickly internalize performance criteria if they are able to use the same over-arching rubric each time a particular skill is measured.
This paper will also explore challenges in the development and validation of such assessments. For example, it is necessary to define 21st century skills in an operational sense, which entails identifying research-based behavioral indicators of these skills one might expect to see in K-5 students. Another challenge is ensuring assessments collect student responses in ways that are developmentally appropriate for young children, who have limited reading and writing capacity. A third challenge is the inability to control factors related to the reliability of tasks. Because these tasks are essentially classroom assessments, the assessment developer has little control over administration and scoring conditions. Indeed, there is an expectation that teachers will modify the tasks to better suit the learning needs of their students. In addition, teachers, rather than trained raters, are scoring their own students’ performance. Thus, generalizability and comparability of scores across tasks and teachers is likely to be limited.
Finally, the paper will share some ideas regarding future trends in PBA. For example, in the coming years, it is likely that the general classification of assessments as either “forced choice” or “performance-based” will gradually be replaced by a system that more accurately reflects a diversity of open-ended item types. In other words, instead of a dichotomy, there is a continuum of assessment types and item formats varying in the extent to which student responses are structured. This continuum ranges from forced-choice items (e.g., multiple-choice, true-false), to highly-structured constructed response items, to so-called “ill-structured” problems. A second trend in PBA is that educators increasingly expect assessments to provide actionable information that can be directly used to improve teaching and learning. This trend suggests that the next generation of assessments, including PBAs, should be capable of providing information that can inform teachers’ next instructional steps.

Author

• Emily R. Lai, Pearson