State-Funded Pre-Kindergarten: What the Evidence Shows. 3. States are beginning to invest in tracking the performance of children in pre-kindergarten

12/01/2003

Reports describing the performance of children who attend state-funded pre-kindergarten are available from several states. Unlike randomized experiments or quasi-experiments, these studies do not have designs that are appropriate for comparing outcomes for children who attend state-funded pre-kindergarten with other children. Instead, the purpose is to follow the progress of one or more cohorts of children who have attended the state-funded pre-kindergarten program. Connecticut is an example of a state that has reported on the progress of children, as well as on teachers and classrooms.(51) The Connecticut School Readiness Initiative (CSRI), a partnership between the State Departments of Education and Social Services, seeks to increase the availability of high quality full-day, full-year child care programs for low-income families and to help bridge the school readiness gap between urban students (primarily minority) and their more affluent suburban peers. The initiative primarily targets low-income three to five-year-olds and provides funding for up to two years of services. Local school readiness councils allocate funding to individual programs.

One local descriptive study of classroom quality in South Central Connecticut was used in part to target quality improvement funds. The results showed that between 1997 and 2000, there was significant improvement in classroom quality, with the number of classrooms rated excellent tripling and the percentage of classrooms rated inadequate to minimal dropping from 50% to 8%.(52)

To evaluate the state’s school readiness initiative more broadly, Connecticut is following two groups of children who attended the state-funded pre-kindergarten across the state in 1998 and 1999.(53) The study assessed pre-kindergarten classroom quality and teacher interactions using standardized and technically sound observations of classrooms. However, the study did not measure the presence of specific interactions known to promote language, pre-reading, and pre-mathematics knowledge and skills. Children’s social skills and very basic school readiness skills were measured using mostly technically sound observations and direct assessments of school readiness.

Results showed that:

  • all classrooms offered care of either minimal or good quality, with 58% being rated as good or better and 2% rated excellent.
  • teacher sensitivity and responsiveness to children’s needs and interests were above average and teachers were neither harsh toward nor detached from children in their care.
  • school readiness skills at the end of pre-kindergarten were close to national norms and the amount of change from fall to spring for children who attended pre-kindergarten in 1999-2000 was the same as national norms.(54)
  • teachers also reported social skills and behavior problems that were similar to national norms.(55)

Other states, such as Delaware, Georgia, Maryland, and Illinois, have collected data that include but are not limited to: children’s performance on state-wide achievement tests, grades, grade promotion and retention, behavior problems, special education referrals and enrollment, teacher ratings of children’s readiness for kindergarten and readiness for each elementary school grade, and teachers’ perceptions of which types of early childhood experiences (e.g., Head Start, pre-K, home care, child care) best prepare children for school. Some states have collected teachers’ classroom observations and compiled portfolios of children’s accomplishments to gauge and exemplify children’s progress. In addition to measuring children’s performance, a few states track the quality of classrooms and teachers.(56)

Enhancements are needed in future efforts to track children’s performance.

Collecting data on the children who attend state-funded pre-kindergarten programs and documenting how children from diverse early childhood settings perform in school are useful for planning new initiatives and improving existing ones. However, it is important to be aware of the technical limitations of the data that need to be addressed in future efforts to track children’s performance and the types of questions that this type of study design cannot answer. For example:

  • Inferences about the effectiveness of state-funded pre-kindergarten should not be made using studies that track the performance of children who attended the programs. Likewise, studies tracking the performance of children from different early childhood settings, such as state-funded pre-kindergarten, center-based child care, home care, or Head Start, should not be used to infer that one early childhood setting produced better outcomes than the other. These inferences can be made only if rigorous quasi-experiments or randomized experimental designs are used to rule out other factors that are also known to affect children’s school readiness and school achievement and that may cause children from different early childhood settings to perform differently in school.
  • Most studies that track children’s progress have the technical limitations mentioned earlier for existing evaluation studies that reduce the usefulness of their results. These limitations should be remedied in future studies.
  • One of the most important limitations of existing studies is the selection of measures. Measures were often newly created or existing measures were modified for a particular study, with no evidence available about their technical soundness. In future studies, new and modified measures should be developed following procedures that allow for fully documenting technical soundness. In addition, measures are often used inappropriately because they were developed for another purpose. For example, “classroom-based” assessments developed for teachers to observe each child’s progress and tailor instruction are often used to track the progress of large groups of children or to make group comparisons. Yet, classroom-based assessments do not have technical properties that allow them to be used for these purposes. Future studies should include measures other than classroom-based assessments that are appropriate for tracking the performance of large groups of children, and making comparisons across groups and to national and state averages.

View full report

Preview
Download

"paper.pdf" (pdf, 782.83Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®