Using Scientific Knowledge to Inform Preschool Assessment: Making the Case for “Empirical Validity”

Abstract

The No Child Left Behind Act of 2001 crystallized the concern for accountability in education. National testing was mandated as a way to improve the “broken” educational system. Publicly funded early education programs were not spared from such testing. While the positive effects of high-quality early education on children’s later school achievement is well demonstrated, too many early care and educational settings in the United States are of minimal or poor quality. Accountability is clearly important for increasing the quality of our early childhood programs, however, it is not yet evident how best to formulate a standard of accountability that reflects the body of knowledge we have gained concerning how young children learn. In this report, we propose two major thrusts designed to bring about a more scientifically informed accountability system: reconceptualizing the ways in which we think about the validity of our test instruments, and reconceptualizing markers of development from products of learning (performance standards) to processes of learning. We introduce the term “empirical validity” to draw attention to the fact that assessments should be built on current empirical work in the various developmental domains. This report focuses on the domains of language and literacy, two areas of major concern for the Federal Head Start program and for which there is an abundance of current research. This body of knowledge provides many examples illustrating how an emphasis on process rather than product can be vital for improving the quality of education. For example, although vocabulary is centrally important and psychometrically adequate tests of early vocabulary exist, these tests do not measure essential aspects of word learning that have been identified as predictive of later language and reading success in early language learning literature. Our case study of language and literacy illustrates how today’s developmental science offers a new knowledge that can be strategically incorporated in assessments for empirically valid testing of children’s competencies. The same argument for “empirically valid” and evidence-based assessments applies to other domains of cognitive growth and to socio-emotional development. The future of preschool assessment would be well served by attention to primary research that focuses on the processes of learning. In this report, we also suggest that one possible avenue for progress in assessment would center on integrative and dynamic assessment techniques that would comprehensively capture the nature of children’s learning, minimize validity concerns related to context and culture, and evaluate how competencies in different developmental domains interact for optimal learning. To bridge the gap between science and policy, developmental scientists and test developers are urged to work together to create innovative ways to chart the developmental processes that support learning and progress toward social maturity in ways designed to ensure that research findings are continuously reflected in current assessments.