Over the past several years, parents with children in day care have been heartened by headlines announcing that out-of-home child care does not appear to damage children's cognitive development or their social and emotional skills.

The headlines stem from findings of the National Institute of Child Health and Human Development (NICHD) Study of Early Child Care, the largest and, many say, most comprehensive longitudinal study of child care conducted in the United States.

But although researchers involved in the study believe they've found evidence that most children in day care don't suffer from the experience, they're cautious not to make firm conclusions about child care's long-term effects.

"The study publications to date are a good first step in answering questions about the effect of child care on child development," says one of the study's principal investigators, Deborah Lowe Vandell, PhD, of the University of Wisconsin at Madison. "But Phase One only takes us to age 3 and to be generalizing from those findings to even what we'll be finding in kindergarten would be presumptuous."

Indeed, as NICHD deposited the massive data set from Phase One onto the Internet in January--it's now available for analysis by interested researchers outside the study--study researchers were busy analyzing data from Phase Two, which examines the children at age 4-and-a-half and in kindergarten and first grade. And they're gearing up for Phase Three, which will examine the children in grades two, three, four and five.

Looking at outcomes in children at these older ages will be critical to any final conclusions researchers can draw from the study. Often, possible links between child care variables such as quality of care and developmental issues such as behavior problems don't appear until children face the higher demands put on them by the school setting. In addition, measures of social and emotional adjustment have been better tested and standardized and therefore are more accurate for older children.

"It will be important for us to look at Phase Two and Three data to see if our findings hold up or how the impact of child care may change over time," says Vandell.

For researchers outside the study, there's plenty to look at within the Phase One data that the study researchers haven't had time or funding to examine. Not only do the data present a longitudinal look at child care, they also include general measures of child development.

Says Sarah Friedman, PhD, the NICHD project scientist and scientific coordinator for the study: "This study is unique in the sense that we characterize in great detail the family environment and out-of-home environment and relate that to a panorama of aspects of child development."

Adds developmental psychologist Sharon Ramey, PhD: "This study is a landmark in the field of child and family development. It has better mapped out the multiple experiences and outcomes for children than any other study."

Assessing the results

A main impetus for funding the NICHD study more than 10 years ago was the long-standing debate in the United States over whether children were better off being cared for by their mothers rather than being placed in out-of-home care for all or part of the work week.

Some early studies--as well as common "wisdom"--suggested the answer was "yes" and that by putting children in day care parents were relinquishing their influence on their offspring's development.

In fact, probably the most important finding to emerge from the NICHD study, say researchers, is that the family environment--including family income, mother-child interaction and mother's symptoms of depression--is far more strongly linked to a child's development than is child care.

"If God gave you the choice between putting a child in a well-functioning, well-resourced family but lots of crummy child care or into a poorly functioning, poorly resourced family and lots of good child care, there's no choice," says study researcher Jay Belsky, PhD, of Pennsylvania State University. "You'd have to choose the former.

"With all the debate about child care, I think we lost sight of that reality," he adds. "Any effect child care has, it has in the background of the overwhelming effect of the family."

Indeed, families that use full-time child care have as much influence on their children as families that use no outside child care, says study researcher Margaret Tresch Owen, PhD, of the University of Texas at Dallas.

Adds Friedman: "Parents shouldn't feel they've relinquished control because family impacts children no matter how long they spend in care."

That isn't to say that the study finds no link between child care variables and child development. However, the size of the correlation is surprisingly small compared with that of the family.

When it comes to the small correlation between child care and child outcome, quality seems to be the critical component during a child's first three years of life. High-quality care--as measured by low child-to-adult ratios, small overall numbers of children and the sensitivity, stimulation and warmth of the child-care setting--is related to children performing well on tests of social and emotional development as well as intellectual development, the study finds. Low-quality care, in contrast, is related to poor performance.

So far, the amount of time children spend in nonmaternal care outside the home doesn't appear to be related to child development, in contrast to findings from earlier studies. But some researchers, including Penn State's Belsky, aren't ready to rule it out as a factor that might show its effect as the children age. In fact, the study found that the more time children spent in care, the more likely they were to have behavior problems at age 2. But that effect disappeared at age 3. The researchers will soon be able to say whether the negative effect comes back by the time children enter first grade.

"What we can measure in children's development at age 3 is not the end of the story," says Owen. "A lot is in flux. In addition, we have better measures as children get older--they're more reliable and more stable."

Mulling over measures

Although the NICHD study uses state-of-the-art measures, those for social and emotional development in young children do not have a long enough history to have been standardized the way measures for language and cognitive development have. And this lack of strong, standardized measures is an issue that worries Stanley Greenspan, MD, clinical professor of psychiatry and pediatrics at George Washington University Medical School. Any time you have a weak measure, he says, you can't be sure that findings of no correlation or small correlations occur because there's truly no correlation or because your measure is too weak to catch it.

"If you pick up background viral activity, but you know you have a weak measuring tool, do you say, 'Forget the virus'?" asks Greenspan. "No. You say that even with a weak measuring tool, you picked up something. So you need to keep an epidemic as a possibility until you get a better measuring tool to prove that wrong."

The same goes for the effect of child care on children's social and emotional development, he says. Since the measures for young children poorly predict later outcomes, Greenspan advises researchers to be modest with the conclusions they draw from the NICHD study results to date.

"There are no definitive answers yet," says Greenspan. "And we certainly can't rationalize any [potential answers] away by saying the home is a stronger influence than day care."

To combat the problem of weak measures for social and emotional development constructs, study researchers used multiple measures for the same construct. To measure a child's compliance, for example, they collected mother and caregiver reports, and they directly observed children in the lab, at home and in day care.

"In that way, we tried to triangulate across observations," says Vandell.

With that said, she and other researchers still have lingering concerns. For one, the meaning of many social and emotional behaviors may not become apparent until the child is older. In addition, the same behavior may have a different meaning for different children. Many 2-year-olds may be noncompliant, but what that means for long-term development isn't clear.

"I share [Dr. Greenspan's] note of caution," says Vandell. "That doesn't mean it doesn't make sense to publish the data. But when we go forward with them, it's important for us to be cautious and put the caveats on what we're finding. It's really a stay-tuned message we're giving."

With that said, the findings to date shouldn't be ignored, says Ramey, director of the Civitan International Research Center at the University of Alabama at Birmingham.

"The effects might change, depending on the age you measure outcomes," she says. "But even if we can't find life-long effects, we can't dismiss the effects at age 3. The idea that we might discount findings from an earlier phase would be a mistake."

Not that the study doesn't have limitations, she adds. Probably the biggest is that sample sizes of high-risk children, including those with disabilities and those from low-income households and certain minority groups, are too small for detailed analysis. She hopes to see a group of special population studies done that map directly onto this study.

Moving beyond child care

That's not to say that researchers outside the study won't have plenty of data to analyze now that Phase One data are available on the Internet. There's no limit to the number of questions that can be asked, say those who know the data. Cornell University's Stephen Ceci, PhD, who reviewed the study for NICHD last year, would like to see an independent analysis of the child care portion of the data by a group of uninvested researchers.

"This is a smartly designed study," he says. "But since the data have so many policy implications, it's important to have an independent analysis that could present a unified interpretation of the findings."

Along with issues of child care, there are hundreds of basic developmental questions to ask. The study has data on physical growth, language development, reading development and sibling structure.

Adds Belsky: "We have all this wood and we've only been at work to build one boat--the child-care effects boat. The same wood can be used to build many different boats."

Researchers can access the study data by applying to the Research Triangle Institute, which analyzes and maintains all the data for the study. For more information visit its web site at http://public.rti.org/secc. Those interested need only show that they will use the data for legitimate research purposes.

Further Reading

  • National Institutes of Health, U.S. Department of Health and Human Services. (1998). NICHD's Study of Early Child Care. [Brochure]. NIH Pub. No. 98­4318.

  • NICHD Early Child Care Research Network. (1999). Chronicity of maternal depressive symptoms, maternal sensitivity, and child functioning at 36 months. Developmental Psychology 35, 1297­1310.

  • NICHD Early Child Care Research Network. (1998). Early child care and self-control, compliance and problem behavior at 24 and 36 months. Child Development 69, 1145­1170.