Early Childhood Educational Intervention

Dr. Francis Campbell: Thank you Rita. It takes a long time to describe a long life and obviously I've had one to date. But I want to talk to you about the study because it's been an exciting endeavor to be involved in. I think it's been a real privilege for me to be involved in it, and I want to help you understand, if I can, what difference in early childhood education really does seem to have made. And let's see if I understand how this slide show works. Do the buttons work? Okay. At first I thought you might wonder, given the kind of conference this is and the heavy emphasis on health, why Maternal and Child Health Bureau would have wanted to fund this study. I went back to look for the things I had when I was writing the grant, and I saw that there was always an interest in developmental disabilities within the Bureau, and one of the goals that they stated was that they were interested in furthering high school graduation, and I heard this affirmed even more this morning in the emphasis on the Bureau, and it has an emphasis on early care and education, and also in helping children to develop to their full potential. This is what this study is all about. There's a real need for this. I don't have to tell anybody in this room that America's children are increasingly--well, I have these things out of order, but individuals who grow up in poverty, and poverty is a big problem in this country, are at risk for developmental delays. They're at risk for school failure, which translates into less ability to be economically self-sufficient when you're grown and probably not as well adapted. Minority families tend to be over-represented among the poor, and I went to the Census just to see what's currently happening to minority families in this country, and discovered that poverty is holding fairly steady, but amongst African Americans it's increasing. For many years, the state and federal governments have really tried to emphasize the health of poor children, and made sure that they got the immunizations and those sorts of basic health services that they need. I don't think there's been as much emphasize on worrying about whether or not they had a good early childhood education. The other thing we have to know is that the risk of children growing up with two parents in the home has greatly decreased. The likelihood that children will be living in single-parent homes is much higher than it used to be and particularly that's true amongst the poor. The fact that so many mothers now have to work means that young children are more likely to need out-of-home care from a very early age. So is there a way that we can combine this need for early out-of-home care and the need for a good early educational environment? As it turned out, the Abecedarian Study was in a really good position to address this question. You might ask, “Well, why are you worried about the very young child's education? Aren't children very malleable, aren't they going to learn pretty much just because they're here? You don't have to do a whole lot for very young children.” We now, though, are beginning to believe that the greatest opportunity for learning happens very early, and that if you miss it, you may not be able to catch up. These are some of the reasons why, and because I'm a little worried about running, I'm going to--Rita has given you some of these points, but I just want to say again that the Abecedarian Study came along when nobody really knew what you could do if you started very young with a very good educational program. So that's why the Abecedarian Study is important, I think. Craig used to talk a whole lot about prevention in those early years, and the idea was that there was this increased risk for such children to be actually called mentally retarded, that is to have an IQ of 70 or lower when they got to school, although nobody thought anything was wrong with them earlier. And so the question was whether he could prevent that happening in a group of very poor children. This is an important study because it's a randomized trial, and what that means is it doesn't have the bias built in of only working with those children whose parents said, "Yes, do something for my child." What we did instead was identify a pool of parents who looked as if they would qualify at random, offer one family this early childhood program and the other family went into the control group. We have had very low attrition in the study, and we have followed the children fairly intensively over--I have data through 21 years, we don't have the 30-year stuff yet. These are the original investigators. As Rita told you, three of them were pediatricians who were interested in group care. I just want to acknowledge that I didn't really start this study, and so I don't want to take credit for it. How did you get into it? We had a high-risk index. There were social demographic factors on it such as the family's income, the parental education, whether or not the father was in the home, whether or not there were other individuals in the family who were showing developmental delays. You had to get at least 11 points on this scale, and the lower your income, the more points you got and so forth. Other criteria was you had to be an apparently healthy newborn. Craig did not want to include children in this trial who already had conditions known to be associated with mental retardation, so a Down Syndrome child would not have been admitted to the study. That was because the whole question was, what is it about the environment that makes these apparently healthy children not look so good when they get to school? So we wanted to start with apparently healthy newborns, and they had to live within commuting distance of the Center because that’s where the program was, and they needed to be likely to remain in the area for at least three years because that's how much money he had. One hundred and twenty families were invited to enroll. Once they learned their random assignment, eight families said, "I don't want to do it." You don't have to be in a study if you don't want to, so we lost them. Two babies that were assigned to one group, they were actually assigned to the control group, the authorities came to us and said, "You will admit that child to your childcare center because we think it will die otherwise." These were children from extraordinarily high-risk homes. You don't turn people down when they tell you that. So those children's data do not appear in what I'm going to show you. And one child actually proved not to be a healthy newborn, and so even though all three of those children attended the Center, their data won't appear here. Well, there were four cohorts of children, they were born between 1972 and 1977, and the original sample, as you see there, was 111 babies born to 109 families. How did that happen? One mother had a set of twins. Craig admitted a sibling, I'll never forgive him for doing it, but one of the children is a sibling. Otherwise, all of these are from different families, and you can see they're reasonably balanced as to males and females in the treatment and control group. According to the high-risk index, they were equivalent to start with. This is the study design. As you can see, we had the original randomization process. We start with a treated and control group and the children remain in their group from birth until they enter Kindergarten at age five. Now when did they start coming to this childcare program? Treatment was in full-time childcare. Our pediatricians did not want them to attend earlier than six weeks, so that's the youngest any child could come. The average age is about four months when they started to attend, but it was in the early infancy stage that they began to come to this full-time childcare. There was a huge need and there still is for good childcare for infants. So we didn't have trouble finding families who thought this was a good idea when they heard about it. What was it like? It was very intense. It was full days, five days a week, all year, except for vacations. It was in a university-based setting. Medical care was available on site, remember those three pediatricians. Children came sick or well unless they had the chicken pox. The staff had extraordinarily low turnover, so these children didn't get a lot of exposure to one caregiver and that person moved on. The ratios were far better than the state of North Carolina would have made them be, one-to-three for infants going on up, one-to-four-or-five for toddlers, up to one-to-seven for preschoolers. This is, actually, a picture of one of the infant caregivers, teachers. The curriculum designed by Joe Sparling is called "Learning Games," and it is being reissued right now if anybody's interested, Joe would love to talk to you. But if you had walked into that nursery, you would have thought it was like any other nursery except you might have noticed that rather than talking to each other, the caregivers were spending a lot of time with the children, and what they were doing with those children were intentional learning activities, but they simply looked like what any mother would have done with a child. I don't know if I can--those are suppose to be interactive, but I don't think I can make them work, so let me just tell you that "Learning Game" showing a scarf is usually done when you're trying to change the baby. But you wear a colorful scarf and you point out to him that this is pretty, and that if he reaches for it, you praise him for doing that and talk to him. They spend a great deal of time talking to the children. As the children got older, obviously, the games became more and more sort of didactic in teaching. Like, can you stack blocks? Or you have two spoons and a fork, and you pick up a spoon and you say, "I have a spoon, can you find another one?" Or something like that. But it's very natural interaction with a child if you watched it happen. The curriculum covered these areas that you see: language, pre-literacy activities, fine motor development, social development. We think that language is particularly important and I'll show you why more later. But here are the preschool results. Here's where I come in because I wasn't in the nursery doing those games. I was independently seeing the children when their mother, not their teacher, brought them in. So I saw the control children as much as I saw the children who were coming to the center every day, and the mother was always the person who was in there with me and the baby or her sister or some family member, but the teachers were not in there. So we didn't teach these tests. These are the standardized instruments that were out there to assess children back then, these are what we used during the preschool years. And that is what we found in terms of standardized cognitive developmental score over the years, and what you see there is that everybody kind of drops off at age two. We think that's a test artifact that they switch to a more verbal test at that point. What you also see is there's a big difference between the children, and that the treated children are definitely showing cognitive gains earlier than those in the control group. Well, what happened when the children graduated from Kindergarten? We had three more years of treatment, and I don't want to talk about that too much because eventually I'm going to revert to the two-group model, but at school age, we had four groups. Half of the children in the preschool were randomly assigned to have three years of follow on into public school and the other half didn't get that, and half the children in the control group had the three years in the primary grades and half the control group didn't get that. So you have children who have eight, three, five--I'm sorry, none, three, five, and eight years of intervention. The school age program, there was a home-school resource teacher; she worked much more through the family. She went into the classroom, found out what the child was working on, where the child seemed to need reinforcement, then she went into the home, showed the mother some activities to help the child learn that, encouraged the mother to use them, worked back and forth as the liaison between the home and the school. If a family was getting evicted or didn't have food or whatever was going on that might have prevented them from being able to help the child, she tried to help them deal with that through referrals as needed. Did it work? It definitely worked for reading beautifully, but school age by itself didn't work very well and this picture kind of shows that. The eight-year group clearly does better, but the group who only got three years, even though everybody's learning to read at the same time, they're not doing nearly as well as the children who had the five years previous, and so what you see there, that's a strong preschool effect, no school age effect at all. What about math? And this for the first three years in school, Kindergarten, first grade, second grade; for math you don't see a strong a stair step for math. What you see is that children with or without that school age program did just about as well. The preschool program helped out in math, and again the five years plus the three years really made those children do really well in math. Oh, and by the time they were 15, there are a couple of other things you need to know. Children who had had preschool intervention were far less likely to ever get referred or placed into Special Education, and in Chapel Hill what that meant was if they had an IEP, they didn't have to be in a self-contained classroom, but if they got state funded services, I call that special services, the likelihood of needing that kind of service was greatly reduced if they’d had the preschool program and the likelihood of being retained in grade. For those children without the preschool program, half of them were retained at some point. A much smaller percentage for the kids who did have the preschool, and so that was another benefit that we saw. And then this shows you how we followed them up. As you can see, we followed everybody intensely, saw them every single year through age eight when all treatment ends. The next time we saw them they were 12 years old, the next time we saw them they were 15 years old, that follows the sort of natural breaks that were in place back then before the middle school. When they finished sixth grade, we saw them, when they finished junior high, which was ninth grade, we saw them again, and thanks to the Maternal and Child Health Bureau, we saw them again when they were 21. Am I doing all right on time? Okay. I want to talk to you about the effects on their intellectual development because in some ways that was the bottom line. Remember I said what Craig wanted to do was to see if he could prevent them falling into really subnormal mental development. I showed you all the points that we have data on these children. We have a lot of intellectual test data on them, so I'm going to describe that. We go back at age 21, when I got the data, and looked at it by itself, just to age 21, you could have knocked me over with a feather when I saw that there was a significant IQ difference because it's not very big, but it was statistically significant. I never thought it would be, but there it was. We did revert to the two-group model because our data for intellectual development and I could show you that, but I don't have time, indicate that the school age program really didn’t affect it. The preschool program did, but the school age program didn't hurt, didn't help. So to really look at long-term effects of the preschool program on intellectual development, we went back and forgot about what their school age assignment had been and said, did they have the preschool or didn't they? That is what we learned. As you can tell, there's a lot I could say about this. You can see that the difference, and these are adjusted scores, that means they're predicted, you can see that the effect is huge when they're little and in treatment, it begins to narrow, and by the time they're six and a half, it's looking pretty narrow, and by the time they're eight, it's narrower yet. But what happened? And what I thought was going to happen, because it’s happened in other programs, is that a few years down the road they don't look different at all. If they was peri-preschool, by the third grade they don't look different at all in IQ. I thought our trajectories would cross or come together. They did not. They change, but they change in parallel, and I think that that is not what I thought we'd find. I'm gratified that it worked. I did not think that the first five years would have that long lasting of an effect, but it did. And I think I've told you what that slide shows, so I won't say it over again in words, but we wanted to know what was it about treatment that made that happen? What mediated the effect of that early treatment? One of the things we knew about these children was how they approached the task when I would test them or anybody would test them, give them something to do, and the whole point of an intellectual test is give you something that you haven't encountered before, for the most part, and see how quickly you can do it or whether you can do it. It's kind of a problem-solving thing. First of all, to do that you have to attend to it, you have to want to do it, you have to be oriented to complete such a task. We knew that when they were very little, children in the treated group were more likely to engage the task that a tester tries to get them to do. So we looked in the long term to see whether that seemed to mediate the effects of treatment, and the way you find that out is you test the model in there with treatment effect alone, and then you put your second mediator in there and you see if that reduces the treatment effect long term. The answer for task orientation is, no. It does in the short term when they are very little, but it does not over the long term. So then we had an independent measure of their verbal skills, we put that in there and low. The effect of treatment is almost wholly mediated by verbal skill it turns out. So when I said the teachers in the preschool were spending a lot of time talking to the children, that went somewhere. We put in other things to see if it had a differential effect on girls and boys. Mother's IQ: we knew that we measured the IQ of the mothers. The mother's IQ, independently, predicts the child's IQ. So if you want to you can say that that's a measure of genetics or it's a measure of the home environment that person provides, but let's just say that the ability of the mother has an effect on the ability of the child. It is a main effect. It seems to work the same way whether the child had early treatment or not. We didn't see a differential effect on girls and boys, but they did change differently. We had a gender-by-time interaction, and this illustrates it, and what it says is little girls seem to go up faster in early childhood, they seem to show this drop-off in adolescence more sharply than males do, but in early adolescence, within later adolescence, the males drop off faster than the females, and so that's a fairly complex pattern. I should have told you, which I did not, that the study sample is 98 percent African American. We didn't set out to do that. It was not intended to be a study of African Americans, but in the neighborhood where we live who qualified and was willing to participate. First of all, we didn't have very many really low-income whites in that community, and the few we found, for the most, part didn't want to do it. So this is 98 percent an African American phenomenon that you see here, and there's a lot of theory about African American boys and whether they care about how they do and things like that in adolescence, but I don't have enough here to really test a theory like that. The kind of home environment that the child had in early childhood makes a big difference. There is a main effect for the home environment as well. Again, it seems to work the same way in the treatment and the control group. In other words, early treatment enhances the child's development, but it does not take away the importance of a good mother and a good home environment, they still are important, and I think that's a take-home message. What you might have thought, early on we thought maybe a sit-down, shut up, and do what I tell you kind of maternal attitude seemed to be suppressing children's development, but that didn't hold up over the long haul. Also, we thought perhaps the father not being in the home might show up as a dampener of the child's development, but in our sample it did not, and I should tell you that at birth about 75 percent of these children were in female-headed households. Well, the size of the treatment effect at 21 is not huge. What difference does a treatment effect that small make in the real world? The evidence is that we know that when they were 21 years old, we tested their ability to read and their ability in math, and those with preschool treatment were doing better. That enhanced educational or academic performance appears to be mediated by the enhancement of the early cognitive development. So that's a difference it makes, and that's their reading scores over time to 21, that's their math scores, math drops off, reading doesn't. I can't go there, I don't have time. Okay, here's some of the things that I want you to really take home from this. When they were 21 years old, if they had had preschool treatment, they had attained more years of education, and the most important thing that I think, to me, is that it was significantly associated with an increase likelihood that they would go to a four-year college or university. Thirty-four percent of those with preschool treatment were in some kind of a four-year college, and that compares to 14 percent of the control group. I don't mean that everybody in the treated group was out there getting a Masters; I don't mean everybody in the control group was out there as a dropout. That's not the case, but the likelihood of going to a school of higher education was significantly enhanced by a preschool program. Here's another one: teen parenthood was significantly reduced if they had the preschool program. We didn't think about that when they were born, but it worked out that way. And I'm being told to "hush," so I would tell you that the National Institute for Early Educational Research up at Rutgers University, Dr. Steve Barnett and his student, Lynn Massey, did a cost benefit analysis based on the outcomes that we saw at age 21. They estimate that on average every dollar spent in the Abecedarian Program will save society four dollars. That figure is being challenged, it may be no more than two and a half, but it still is cost effective, and I think that's something that you might want to know too. Why is it cost effective? Because of the likelihood of the fact that they had more education, they projected on average they will earn over $100,000 more over their lifetime. There's actually a benefit for mothers of these children because remember, those women got five years of free, excellent, full-time childcare. That gave them a chance to make some gains of their own, so we estimate that they will do better, and there was a savings in Special Ed and grade retention. And, the final thing I'll say is they were less likely to report that they smoked when they were 21--back to health. So I'm going to let Robin tell you some things that some of the research to practice that we think have happened in North Carolina, not just based on this study but this study has helped to support some initiatives in the state.