Evaluating Children In Preschools Programs
Growing interest in publicly funded programs for young children has drawn attention to whether and how Head Start and other early childhood programs should be asked to prove their worth. Congress asked the National Research Council for guidance on how to identify important outcomes for children from birth to age 5 and how best to assess them in preschools, child care, and other early childhood programs.
The Research Council's new report concludes that well-planned assessments can inform teaching and efforts to improve programs and can contribute to better outcomes for children, but poor assessments or misuse of the results can harm both children and programs. The report offers principles to guide the design, implementation, and use of assessments in early childhood settings.
Federal agencies, states, school systems, and other organizations that evaluate early childhood programs or the children they serve should make the purpose of any assessment explicit and public in advance, the report says. For example, a state should specify whether an assessment will be used to help teachers gauge the progress of individual children or to help public agencies decide whether to continue a program's funding.
"The goal of the assessment should guide the choice of the assessment tools used, and assessments that will have widespread effects should meet high standards of rigor and validity," said Catherine Snow, a professor at the Graduate School of Education at Harvard University and chair of the committee that wrote the report. "For example, using a standardized test with a sample of children in a program would be suitable if the goal was to determine whether the program is bringing children closer to national norms, but if the purpose is to guide instruction within a specific classroom, a nonstandardized assessment linked to the curriculum would be appropriate."
Effective assessment must be part of a larger system with a strong infrastructure to support children's care and education, the report says. Facets of this system should include clearly articulated standards for what children should learn and what constitutes a quality program. Other aspects include professional development opportunities, training to familiarize policymakers, teachers, and administrators with standards and assessments, and continuous monitoring to ensure that all elements of the system are working together to serve the interests of the children.
The report urges extreme caution in basing high-stakes decisions -- such as determining whether a program will receive continued funding or whether a child is eligible for services because of an identified disability -- on assessments of young children. Models such as those set forth in the No Child Left Behind Act strive to link yearly progress assessments to explicitly defined academic content areas for children in grades three through 12. It would be inappropriate to borrow this model unchanged and apply it to early childhood settings, the committee said, because well-defined academic content areas are not characteristic of excellent care and education for younger children.
Cutting a program's funding or imposing other negative consequences based on assessments of the participating children should happen only under certain conditions -- if the program has been given enough resources to meet expectations, for example, and if the level of children's development when they entered the program has been taken into account. Child assessment results should never be the only information considered. And a program should not be closed or restructured if doing so would have worse consequences for children than leaving it open, the report adds.
Likewise, decisions to penalize a teacher should never rest solely on findings from assessments of students in his or her classroom, without considering children's starting points, how the test is related to the curriculum, and whether the teacher has adequate support, professional development, and other resources.
Programs' quality should be evaluated based not only on how they affect children's academic skills such as language and mathematics, but also on whether they improve other important aspects of child development, such as social and emotional skills, the report says. While good measures of certain outcomes -- such as literacy and language development -- currently exist, tools to assess other abilities such as problem-solving and creativity remain underdeveloped, and more effort will be required to improve their quality.
In addition, the report notes, some assessment measures have only been tested with populations that do not represent the diversity of children enrolled in today's early childhood programs. Care should be used in assessing the status or progress of young children with special needs and those for whom English is a second language, because many existing instruments have not demonstrated their validity for these groups.