1. Goals become guides for evaluation. Be sure that evaluation is not so limited that unexpected, but important, information gets overlooked.
2. Evaluation is used for continuous improvement.
Programs conduct regular self-evaluation that focuses on implementation rather than simply outcomes.
3. Comprehensive goals are used.
These include well-defined early learning standards and other program goals in addition to child outcomes, such as those that relate to community, families and staff.
4. Evaluations use valid designs.
Methods need to be scientifically valid and reliable to be effective. The evaluation method should also be aligned to the program’s goals; otherwise, mismatches can lead to mistaken assessment of interventions (Yoshikawa & Zigler, 2000; Muenchow, 2003).
5. Multiple sources of data are available.
Information about the staff, program, and demographics provide the context for evaluating assessment results.
6. Sampling is used when assessing individual children as part of large-scale program evaluation.
This is a way to protect children from “the burden of testing” and minimize the chances that data will be misused.
7. Safeguards are in place if standardized tests are used as part of evaluations.
These include developmentally, linguistically, and culturally appropriate testing. Data quality checks should also take place regularly.
8. Children’s gains over time are emphasized.
A record of the progress that children make, rather than test scores, should be included in program evaluations.
9. Well-trained individuals conduct evaluations.
Effective program evaluation needs to be unbiased to be fair. Only qualified people should formally assess and analyze programs, with input from staff of course. All staff should be assisted to improve their “assessment literacy.”
10. Evaluation results are publicly shared.
All “stakeholders” — parents, policy makers, staff and others — should know the results of program assessments.