RESEARCH LIBRARY

Attempts to measure so-called "non-cognitive" factors for the purposes of educational policy and practice are relatively recent. The authors identify serious challenges to doing so. They discuss advantages and limitations of different measures, in particular: self-report questionnaires, teacher-report questionnaires, and performance tasks. The authors discuss how each measure’s imperfections can affect its suitability for program evaluation, accountability, individual diagnosis, and practice improvement. In addition to urging caution among policymakers and practitioners, they highlight medium-term innovations that may make these measures more suitable for educational purposes.