Monday, June 9, 2014
The podcast is available at http://www.nichd.nih.gov/news/releases/Documents/060914-HeadStart.mp3 (MP3 - 1.2 MB).
Ms. Rebecca Lazeration: Head Start is a program which provides low-income children with preschool education, health care, and nutrition services. A recent analysis of a national study on Head Start shows that 1 year of the program improves children’s math, literacy, and vocabulary skills. However, the program has especially strong effects in certain children.
Elizabeth Miller and her colleagues analyzed data on over 3,000 children from the Head Start Impact Study. For early math skills, Head Start was most helpful in children with the least-enriching home environments; that is, children whose parents spent little time reading to them or counting with them. Children whose homes provided a medium amount of such stimulation made the biggest gains in early literacy skills.
From the National Institutes of Health, I’m Rebecca Lazeration and this is Research Developments, a podcast from the NIH’s Eunice Kennedy Shriver National Institute of Child Health and Human Development, the NICHD.
With me now is Ms. Elizabeth Miller, a Ph.D. candidate at the University of California, Irvine. Ms. Miller, thank you for joining me today.
Ms. Elizabeth Miller: And thank you for having me.
Ms. Lazeration: For your study you analyzed data gathered previously by a national study of Head Start. The evidence shows that Head Start is a very successful program. Why did you want to learn more about the effect of educational environment at home on the benefits of Head Start?
Ms. Miller: Yes. So, one of the primary interests with the original Head Start Impact Study was to determine whether the effects of Head Start really differ for different groups of children and for whom the program worked best. It’s logical to expect that, with a complex program like Head Start, not all children will be affected the same way because a lot depends on the match between what they bring to the table in prior experiences and the Head Start environment.
So, the original Head Start Impact Study examined different sub-groups of children, such as different race/ethnicities, initial skills and abilities, for example, to see if there are any differential program effects. And there was some evidence of these moderated effects. So, for example, dual-language learner children differentially benefitted from assignment to Head Start compared with monolingual English children on some outcomes.
So, given the importance that we know of parents and the ensuing home environment, we really wanted to extend the work of the Head Start Impact Study by examining whether the effects of the program differed by the level of pre-academic stimulation that children received at home. And so that was really the basis for this study.
Ms. Lazeration: Now, you used the term “pre-academic stimulation.” Can you explain a little bit more about what that means?
Ms. Miller: Sure. So, this was a scale that was developed for the Head Start Impact Study [HSIS], as well as with using the Head Start Family and Child Experiences Survey, and it’s loosely based on Caldwell and Bradley’s HOME measure. And it asked parents in an interview about educational activities that they do with their children, such as writing letters and words, practicing with numbers, vocabulary, talking about the calendar, etc. So, the HSIS parent interview asked parents how often they did these activities and if they did them, yes or no, and we combined them into a pre-academic scale.
Ms. Lazeration: Now, overall, what were the effects of pre-academic stimulation on the children’s skills?
Ms. Miller: Right. So, it was interesting. There was a nonlinear main effect pattern for parental pre-academic stimulation on children’s outcomes, at least in the sample of Head Start children that we had.
So, for example, children from homes with lower levels of pre-academic stimulation perform worse than children from homes in the middle ranges, on average, on all three outcomes that we looked at. However, children from homes with higher levels of stimulation really didn’t differ significantly from children from homes in the middle ranges. So, it was essentially like this nonlinear pattern from low to middle and then not much of an increase from middle to high.
Ms. Lazeration: Now, when you broke down your results by skill, you saw some interesting differences, for example, with math. Head Start had a relatively big effect on gains in math skills in kids coming from the least-enriched homes. Do you have an idea of why that would be?
Ms. Miller: So, we thought a lot about this because, obviously, when you see different patterns emerge, you’re curious as to why. So, a lot of previous research has actually found what we called evidence of the compensatory effect when it comes to children’s early math skills. So, for example, children who come from homes where they aren’t doing a lot of math skills, in the past had been seen to have the biggest boost from a program like Head Start or it could be other childcare experiences. And that was similar to what we’ve seen here.
And one reason hypothesized why is, really, that parents in general don’t do much math with their children and so it may be because you’re going from an environment where you’re not really doing a lot of math at all to one where math is definitely part of the curriculum so you see these bigger boosts, especially if parents really are not doing a lot of early math stuff at home. So, it was really consistent with what prior research has found in this domain of early math skills.
Ms. Lazeration: Can you give some examples of what early math skills are in preschoolers?
Ms. Miller: Right. So, we essentially used the Woodcock-Johnson Applied Problems test, which is a well-known achievement test that has been used in preschool age. And so, they’ll ask some questions about pattern recognitions, about shapes—basically, the building blocks of early math skills that, in subsequent years, children will build on those skills.
Ms. Lazeration: And then for early literacy, the effect was a little bit different than on math. Can you explain the difference of what you found and why you think that that happened?
Ms. Miller: Sure. So, it was very interesting. For the domain of early literacy, we found what we called a Goldilocks effect, in that children from homes in these middle ranges received the largest boost in score from assignment to Head Start compared with either the higher or lower ends.
Now, this is really interesting because it might imply that there’s a sweet spot of sensitivity in which children who receive just enough stimulation at home would benefit the most from a program like Head Start. So, not too little so the program can properly build on the skills that are taught at home and not too much so that it’s not just a repetition of what they’re doing at home but, really, these middle ranges.
And what’s interesting is in other previous work with low-income populations, not having to do with the home environment, but there have been some studies that have examined welfare-to-work policies, which found a similar relationship in that mothers who were moderately hard to employ had a lot more success with the program than those who were easily employable and not that employable at all.
In developmental science, a lot is about the match between what the person has in their skills and their dispositions and then the environment that they’re encountering. And so, it might be that, for a domain like early reading or early literacy, we’re really seeing that kind of pattern that the middle ranges receive the largest boost.
Ms. Lazeration: And can you explain what early literacy would look like in preschoolers?
Ms. Miller: Sure, OK. So, for this we used the Woodcock-Johnson Letter-Word Identification test, which was from the same battery of assessments as Applied Problems. And so this, for example, will show children four letters and it will say, “Can you identify which one is the D?” and they’re supposed to point to the D. And then it gets more complicated from there. It will eventually go to whole words and even more complicated with older ages.
Ms. Lazeration: Now, do you have any ideas about why the results of these two skills had such different patterns?
Ms. Miller: Yes. So, again, that was something that we thought a lot about and the truth is we went back to the original HSIS report and we found evidence of these different patterns of moderation with other sub-groups. So, for example, we saw that Head Start impacts on letter-word and applied problems really tended to vary in non-monotonic ways. So, for example, you had children of mildly depressed mothers; the impacts were larger for them than children with parents with either no depressive symptoms or very severe symptoms.
So, the reason why we really wanted to explicitly test for these three hypotheses is because we had seen some evidence in the final report that there were different patterns and we really wanted to know if it extended to the home environment as well.
Ms. Lazeration: And what should preschool programs and parents of preschoolers be taking away from the research that you found?
Ms. Miller: Sure. So, first of all, and I think first and foremost, we found evidence in our study of the critical role that parents play in children’s early learning. So, in our study, even if children were assigned to the control group, if parents provided a lot of pre-academic stimulation at home, their children really performed similarly to Head Start children. And Head Start is a program that really values parents and treats them as vital partners in their children’s development and our study really enforced that. So, that’s, I would say, first and foremost.
But, second of all, with a program like Head Start, we really found no support for the accumulated advantages hypothesis, which essentially says that those children who come to programs with higher level of skills or with less-risky home environments might benefit more from assignment to Head Start. But our study really found the opposite and we really believe that Head Start should continue to serve children at higher risk as those of more moderate risk because this match between the childcare environment and with the home environment is really crucial. And it seems to be even more crucial for children who really have some element of risk, I would say.
Ms. Lazeration: Ms. Miller, thank you so much for speaking to me today.
Ms. Miller: Thank you. I appreciate it and I’m glad people are excited about the study.