As the federal government’s Head Start Bureau proceeds with a new testing program for 4- and 5-year-olds, criticism of the test—called the National Reporting System—continues to mount among experts in early-childhood education.
This month’s edition of Young Children—the journal of the National Association for the Education of Young Children—includes an article that calls items on the test “rife with class prejudice” and “developmentally inappropriate.”
“The idea that a narrow test of young children’s skills in literacy and math can represent a quality indicator of a holistic program like Head Start shows a stunning lack of appreciation for the comprehensive goals of the 38-year-old program,” write Sally Atkins-Burnett, an assistant professor of early childhood and special education at the University of Toledo, and Samuel L. Meisels, the president of the Chicago-based Erikson Institute, a graduate school in child development.
The test is actually a series of subtests drawn from other assessments that fulfills a mandate included in the last congressional reauthorization of the federal preschool program in 1998. It is intended to help officials better determine how children are progressing, to improve teacher training and technical assistance, and to be used in monitoring local Head Start grantees.
The test was implemented this past fall with all English- and Spanish-speaking children—about 500,000—and is scheduled to be given again this spring.
‘Far From Consensus’
A $6.5 billion program that serves more than 900,000 poor children, Head Start is currently up for renewal by Congress. While members of both the House and the Senate introduced legislation last year to stop the National Reporting System until further research could be done, the testing program itself is not part of the current reauthorization process.
But growing concerns about the validity of the results have some observers urging the Bush administration to slow down and revise the assessment program.
And while Mr. Meisels—who also serves as a member of a technical work group for the test—was one of the earliest and most vocal opponents, he doesn’t appear to be the only expert with doubts about the usefulness of the test.
Observers who attended a Dec. 17-18 meeting in Washington of the 16-member work group said that other members of the group have joined Mr. Meisels in suggesting that only a sample of Head Start children be tested instead of every child in the program. In addition, Mr. Meisels wants the results to be used only for internal purposes, not to hold programs accountable.
The members of the work group “are far from consensus” on the purpose of the test, said Adele Robinson, the director of public policy and communications for the NAEYC. She attended the meeting.
For instance, an item on the test that asks children to identify the facial expression “horrified” has been cited by critics as one example of how the questions can be confusing. Four Caucasian faces are pictured, and the children are asked to pick the one that most looks horrified.
That item ignores “the fact that facial expressions differ in different cultural and ethnic groups,” Ms. Atkins-Burnett and Mr. Meisels write. They add that at age 4 or 5, children might be too young to distinguish a horrified look from one of anger or rage.
Test-Administration Errors
Findings were also presented at the December gathering showing errors in the way some test administrators had asked children test questions or scored their answers. Critics say those mistakes provide one more reason that Head Start officials should hold off on using the results to make decisions about grantees.
“We can either go by this timetable,” Ms. Robinson said, “or we can wait for a consensus from the experts.”
But Steve Barbour, a spokesman for the Administration for Children and Families, which oversees Head Start and is within the U.S. Department of Health and Human Services, said a halt in the next round of testing was unlikely unless members of Congress acted to do so.
“The horse is already out of the barn,” he said, noting that the data from the fall testing period were already being analyzed and that results could be released by March.
He also said that the questions used for the test were taken from reliable measures that have long been used in the field.
