69传媒

Questions over the long-term effectiveness of Head Start鈥攁nd debates over the research into it鈥攁re almost as old the program itself.

As early as 1969, an educational testing and data-reporting organization called the Westinghouse Learning Corp. conducted the first major evaluation of the program, finding that the cognitive and language gains seen among Head Start participants at 1st grade had disappeared by 2nd or 3rd grade.

But critics argued with the study鈥檚 design, saying the comparison group may not have been as disadvantaged as the children in Head Start.

A similar debate took place in 1985, after the U.S. Department of Health and Human Services released the results of a sweeping review of research showing that immediate gains in health and in cognitive and social/emotional skills dissipated over time. The authors wrote that the 鈥渢est scores of former Head Start students do not remain superior to those of disadvantaged children who did not attend Head Start.鈥�

Those conclusions were also criticized, largely because of the sheer number of the studies analyzed鈥攎ore than 200, which, some observers said, could have caused positive effects on children to be minimized.

Still, opponents of big government programs have continued to play up the notion of 鈥渇ade-out.鈥�

鈥淚f students test the same with or without Head Start after a year or two, what鈥檚 the point of sending them through the program in the first place?鈥� Darcy Olsen, the president of the Phoenix-based Goldwater Institute, wrote in 2000 when she was the director of education and child policy at the Cato Institute in Washington.

Some Head Start supporters have noted that many studies of effectiveness only consider academic gains and don鈥檛 measure the benefits of the comprehensive health, nutrition, and other social services provided for families.

The Head Start Family and Child Experiences Survey鈥攁n ongoing project following three cohorts of Head Start students鈥攈as gathered data on both social and academic skills. A recently released follow-up of the 2003 cohort showed that they improved in cooperative behavior and were less withdrawn at the end of their kindergarten year. They also made gains in early reading, vocabulary, early math, and early writing.

But while the early-reading skills of those children reached the national norm, they were still well below the national norms in vocabulary and math, the survey found.

More Study Sought

Policymakers have long called for a randomized study of the program. But Head Start advocates opposed such a design on the grounds that they couldn鈥檛 support a project that would intentionally deny services to poor children.

Even so, the Head Start Impact Study鈥攊n which children who attended the program are being compared with those who did not鈥攂egan in 2002 and is continuing. Its control group is made up of children who could not get into the program because all the slots were filled after a lottery, explained Nicholas Zill, the director of the Child and Family Study Area at Westat, a Rockville, Md.-based research organization.

Initial results released in 2005 showed 鈥渕odest鈥� gains for the Head Start children in pre-reading, pre-writing, and vocabulary skills. But improvements were not found in oral-comprehension or math skills. Results after the children鈥檚 kindergarten year are being analyzed and will be released later this year.

Officials with the Bush administration noted that the preliminary findings showed that children in the program still lag behind their peers, while Head Start advocates used the results to boast that the children are making progress.

鈥淭he politicians have a very different impression of what the program is supposed to do, and what it can do,鈥� Mr. Zill said.

In a new paper, two researchers at Georgetown University also suggest that what appear to be minor positive effects on children who attended Head Start could still mean that the program is cost-effective in the long run.

Jens Ludwig, a public-policy professor, and Deborah A. Phillips, a psychology professor, write that 鈥渟hort-term boosts in academic skills鈥� can be viewed as a 鈥減roxy for the bundle of early skills that promote long-term outcomes,鈥� such as motivation and persistence.

Still, Mr. Zill said that while children in Head Start make learning gains, they are 鈥渨ell behind middle-class kids鈥� when they start the program, and that 鈥渕aking modest improvements doesn鈥檛 close the gap in a major way.鈥�