Collaborative for Authentic Assessment and Learning

Overview

In our current educational culture, standardized testing is often seen as the only reliable yardstick for measuring student ability and knowledge. Extensive research shows, however, that standardized testing does not itself promote deep learning, nor does it measure our most valued educational outcomes (such as problem solving or teamwork). To better measure how education supports student ability and knowledge, and therefore captures information needed to improve educational practices, we need to develop and spread valid alternatives to standardized testing.

In September 2009, AAC&U, having partnered with over 100 faculty from colleges and universities around the country, released fifteen rubrics designed to evaluate real-world student work samples in a variety of important skill and knowledge areas. The practices enabled by these rubrics form the beginning of an alternative to standardized testing: evaluating actual student work in order to determine student skill and knowledge levels.

In order to enable the collection, analysis, and dissemination of this alternative evaluation methodology within and across adopting institutions, AAC&U now proposes to develop the Collaborative for Authentic Assessment and Learning (CAAL). CAAL is an extension of the original VALUE (Valid Assessment of Learning in Undergraduate Education) project that developed the fifteen VALUE rubrics.

The Collaborative for Authentic Assessment and Learning (CAAL) seeks to legitimate the evaluation of actual student work in real educational contexts as a measure of educational program effectiveness for contexts like accreditation, as an alternative to the practice of standardized testing. Over time, CAAL would also seek to establish benchmark data on student performance in different educational contexts.

To achieve these goals, CAAL will seek funding for and build, test, and release a centralized repository and a data collection/access protocol or standard for the repository. The centralized repository will house score data from actual student work evaluated against validated rubrics--originally the Association of American Colleges & Universities (AAC&U) VALUE rubrics and possibly, in later phases of
the project, other rubrics. The repository would be designed to accept rubric score data directly from any educational software platform following established protocols or standards. The repository would also be designed to allow educational institutions without access to a compatible educational software platform to upload rubric score data manually from spreadsheets or other standard data formats.

The centralized repository will seek to publish fully public reports of anonymized, aggregated data, benchmarking student performance across multiple institutions as a measure of program effectiveness. These reports would be much like the reports educational institutions receive from the National Survey of Student Engagement (NSSE). Otherwise, data in the repository would be private.