The distinctive role of administrators in the assessment of young children is summarized by Hills (1992):
"Unlike teachers and parents, administrators do not need detailed information on individual children. Instead, they need to know how groups of children are doing in relation to a standard of expectation. Administrators are accountability agents; they must monitor how programs fulfill their goals and objectives and assess needs for broad program changes.
Currently school administrators are required to give significant time and energy to large-scale program evaluation, often implemented through formal, standardized testing programs beginning by Grade 2 or 3. Pressure for high scores on standardized tests--reinforced by media reports, politicians, government officials, and educational reformers . . . weighs heavily on school superintendents and principals, who are accountable to the public for the outcome of publicly funded programs. . . . Administrators have an equally important responsibility to be instructional leaders, supporting the kind of assessment that teachers need to improve the quality of children's experiences. Too often administrators lack experience and training in early childhood education. When these dual roles conflict, administrators should seek in-service training on young children's development and learning, curriculum and assessment, and teachers' roles in early childhood education. Administrators so informed may be empowered to help teachers observe and evaluate children's performance and reduce unexamined reliance on mass standardized testing for program evaluation and teacher assessment." (p. 47) (see also Stiggins, 1985, 1991)
Meisels (1993) believes that appropriate assessment of young children makes demands not only on teachers but also on policymakers and administrators. He notes that while in past decades, educational administrators and the policymakers they advise could rely on "simple quantitative data that have been used on countless occasions," the responsibilities are now different; administrators must now understand that "achievements must replace achievement scores, observations must replace inferences, and assessing to find out what children know and can do must replace testing to find out what children don't know" (p. 39).
According to Shepard (1989), "Policymakers must also understand and preserve the important distinction between classroom assessment and accountability testing" (p. 7).
The implications of these administrative responsibilities are that administrators must support and promote professional development opportunities that help teachers gain understanding and skills in appropriate assessment, work for adequate supervisory support for teachers' attempts to implement it, and communicate well and positively to policymakers and parents the importance of appropriate alternative assessment.