20 April 2010

Early childhood experiences and school achievement: Do trajectories start earlier than we think?

Linda J. Harrison
Associate Professor of Early Childhood Education, Charles Sturt University

Abstract and audio of presentation

Edited transcript

The following audio presentation is brought to you by the Australian Institute of Family Studies as part of our monthly seminar series in which we showcase national and international research related to the family.

The seminars are designed to promote a forum for discussion and debate. They are open to the public and free of charge.

Seminar facilitated & speaker introduced by Dr Jacqui Harvey.

Linda Harrison:

Thank you, Jacqui, and thank you all very much for coming. I've had a chance to look at the list of names and there's such a huge variation of people here, so I imagine there'll be lots of very different and interesting questions at the end. What I'm going to try and do is make sure that if things need to be rushed through I can do that just so that we've got time at the end for some nice questions about where to next. A lot of this is the "where to next?".

The topic has a question - "Do trajectories start earlier than we think in terms of young children's development?" I raise that question in a very specific context, where I pose that question in the context of active and exciting changes in early childhood education and care.

As many of you would be aware, there's a very forward thinking government in place in terms of early childhood education and care and they've made some major changes and inroads into that.

In fact, I was just thinking the other day how much more successful the COAG process was in terms of early childhood than it is currently around the health initiatives. In fact, it was so successful there was hardly a murmur really going on about it and some of you probably enjoyed that yourselves or weren't even aware that it was happening.

What I want to do today is focus on those new directions that the government has moved us towards in relation to early childhood education and care. (slide 2)

The National Early Childhood Development strategy is an important document that covers a lot of different areas of child and family life (and I've brought materials along with me today so if anyone wants to have a look at those afterwards or just familiarise yourselves with some of the material I'm referring to, I have the hard copy and also will show you or give you some website links at the end).

These are some of the key directions that are setting this context; and it's a context of change. These things only just came into play last year, so if you look at the dates we've got July, then December 2009; and some of these policies are being rolled out this year. So a lot of change for preschools and childcare services (long day care and family day care), which some of you may be involved in yourselves.

Just to mention what some of these are - the first direction that the government decided to go in was to develop the Early Years Learning Framework which is an overarching curriculum framework for all early childhood facilities in the country. It was adopted and has been agreed to by all of the States/Territories. This means that there will be gradual changes in terms of staffing and regulations and also in terms of State-based curriculum documents. Some will change the curriculum framework for the Early Years Learning Framework (EYLF); some may want to work with the EYLF and their existing state or territory curriculum documents.

But basically that new document (EYLF) has been embraced by the field. That's my feeling having worked on the preparation of that document. It was a Charles Sturt University led initiative that won the tender to do that work which started in 2008 and was finished in July of 2009.

The other major initiative has been to develop a National Quality Standard that moves forward from the existing National Childcare Accreditation system of quality assurance (which is only for childcare services). The National Quality Standard will apply to all under-five settings and also school-age care.

Again, this is a major shift which is bringing together what has been, and in some cases still is, a fairly traditional divide between childcare and preschool programs.

Third, is the government's promise to provide universal access to a quality preschool program for all four to five year olds. In other words, to provide preschool for the children in the year before they start school.

Interestingly you'll see from the LSAC data that I'll present later that most children seem to attend a prior to school program, which includes long day care and preschool programs. When you only look at something that's just called "preschool" then the numbers look a bit lower, but when you include centre- based childcare then there's a pretty high take up in the year before children start school.

The difference being proposed by the new government policy is that these programs will all be taught or overseen by someone with a university qualification. There will be 15 hours of preschool provided free to the family for the child, provided by an early childhood qualified, university qualified teacher.

So, as a result of that, another area that is needing reform is a workforce strategy, because early childhood services are characterised by two aspects that have been very slow to change. One are staff shortages of degree qualified staff, and those of us at Charles Sturt and in other universities are working very hard to find ways to increase the uptake of university qualifications in early childhood.

Second are problems with staff turnover that have been a problem for the sector for a very long time and will probably continue to be so.

That's just by way of introduction.

What these reforms are aiming for, that is the overall goal and the vision that COAG has agreed on, is that "by 2020 all children have the best start in life to create a better future for themselves and for the nation".(slide 3)

Again, I've pulled out some quotes that are specific to the early childhood aspects of the National Strategy. It's reassuring to see that there's a recognition and valuing of children for who they are in that first quote, but also who they're going to be. If you read the title of the Early Years Learning Framework you can see we've highlighted not just the "Becoming" part but the "Belonging" and the "Being", so who children are when they come into early childhood services.

The early childhood reforms are designed to activate or set things in train to achieve this vision; and really it's a very big ask of the early childhood sector. The vision is based on evidence and the back section of the National Strategy document is devoted to five or six pages of evidence. The early childhood evidence is primarily based on these two studies that I've summarised here (slide 4); early intervention studies that have been conducted in the United States and the more recent UK study, The Effect of Provision of Preschool Education.

I want to summarise some of the key findings from these for a couple of minutes, to tell you what these strategies have been based on and, in a sense, where the government wants early childhood to go and what achievements they expect to see from these sorts of initiatives.

The studies are very different. The targeted intervention programs are based on findings from 30 to 40 years of research. The Perry Preschool Program started in the 1960s and is something I've been teaching my students about for a very long time. It's been highly successful, or shown to be highly successful, for a particular group and also for a particular type of program. This is where the figure that some of you would be aware of, seven dollars being saved for every dollar spent, comes from. A very important document that James Heckman produced that showed the economic benefits of early childhood education and care is strongly based on these data.

More recently, though, is a really important study conducted in the UK, The Effect of Provision of Preschool Education, or EPPE for short, which is not a targeted program. It's universal in the sense that they recruited children from wherever they were getting their preschool program. That included, in the UK system, nursery schools, playgroups, integrated centres, a whole range of them across the country and a range of different local education authorities. They followed these children into their early years of school.

What's unique about this study, however, and also important to keep in mind, is that it started at age three. It's a preschool-based program. It's not looking at what happened prior to the age of three.

EPPE also had a sample of children who were not attending any programs, their control group. You can see the numbers in this study are much healthier and larger for quantitative analyses and comparisons than the Perry Preschool Program: 3000 children who were attending different types of programs.

(Slide 5) This is a slide that I was given and allowed to share by Ted Melhuish who presented the results of EPPE at a conference in Melbourne several years ago. It summarises the key findings about the quality of the programs. Much of the information, or much of the grounding, behind decisions to put money into achieving higher quality early childhood programs is coming from this kind of evidence.

What you can see they've done here is to quantify the effect of going to preschool or having preschool education in terms of the equivalent benefit in the literacy tests that were conducted at age five. For children who attended preschool for up to two years, all of the children had some sort of benefit compared to those children who didn't attend a preschool program at all.

But the benefit increased as the amount of time increased. So the best outcome was for children who were in high quality programs for two to three years who had up to seven or eight months benefit according to the effect size that was calculated.

Slide 6 shows another key finding. This is the evidence for arguing that universal programs and high quality programs make a difference for children growing up in disadvantaged backgrounds, shown by the difference between the dotted line and the solid line for those children who did and did not attend a preschool program. Again, this is referring to any kind of preschool program.

What this slide is showing is that, for the children on the right hand side in the group from semi-skilled and unskilled families, who are the ones in more disadvantaged areas and the children one would expect preschool to make a difference for, it does make a difference. Those children who attended preschool were higher by enough points to be above the expected minimum level for Year 2 reading.

But what I find interesting about this slide is that the two lines are pretty well parallel, showing that everybody is benefiting by the same amount. Remember, this was not a targeted program, the EPPE program. It was just your normal kind of early childhood preschool activities, and with some centres doing better than others, so quite different in that way from the U.S. research.

Coming back then to where we are here, the points (slide 7) pull out some of the key areas that I've identified in the National Standard and the sorts of policies that the government is expecting to put into place here and the sort of outcomes that they're expecting to see, from my reading of it.

The focus very much is on learning outcomes. The Early Years Learning Framework contains a large section on learning outcomes. That's different from a focus on standards (not that I want to go into that) but the aim is to improve the outcomes for children. So it's a more direct approach in terms of the changes that are expected.

The outcomes and learning competencies are focused on early literacy and numeracy competencies. This quote is taken from the Melbourne Declaration on Educational Goals for Young Australians identifying essential foundations for success in all learning areas -  early literacy and numeracy is a critical aspect of that.

Now let's look at some of the evidence for this and for whether or not early childhood contributes to that.

First of all, before coming to the early childhood question, I want to review the evidence for the statement that early literacy and numeracy sets children on track for school success.

So I'm drawing on two pieces of evidence here. One is an international study that was published in 2005 by Duncan and his colleagues, one of whom included Amy Claessens whose name appears there for the second piece of evidence.(slide 8)

They reviewed studies from several different countries, not just US studies, and showed very clearly that early academic competence at the start of school in these three areas, maths or literacy or attentiveness, attention in class, set the foundation for school performance in the middle and upper years of primary school.

So in other words they were thinking about trajectories. This evidence is showing very clearly that where children are at when they start school at age four to five sets things in train for their future school success.

Amy Claessens, who was part of that international study of six studies, was in Australia last year (and some of you may have heard her speak), where she presented on the LSAC study. She did a very similar analysis using data from the Longitudinal Study of Australian Children.

I've just summarised the findings that she presented here and her full talk (in PowerPoint) is available on the LSAC conference website if you want to read that.

Basically when she looked at teachers' ratings of children's literacy and mathematical abilities at age six to seven and again at age eight to nine, she found that both of these outcomes were predicted quite strongly by children's school readiness skills, which included a measure of receptive vocabulary - the Peabody Picture Vocabulary Test, and also early literacy and numeracy tested by the Australian measure, the Who Am I? test, and also ratings of attentiveness. So she included the three areas that had been identified in the international review.

The main predictor was the literacy and numeracy test, the Who Am I?. Results for the Who Am I? were the strongest predictors of children's school achievement at age six to seven and eight to nine.

Before coming into what I want present, which is about the early childhood component of this equation, I just want to give you a very quick summary of the Longitudinal Study of Australian Children. Many of you would be very familiar with this so it's very, very quick. (slide 9)

Basically there are 10,000 families around the country who were recruited in the first wave in 2004, half of whom had a baby aged six to 12 months and half of whom had a four to five year old, on average four and a half to five years of age.

These families are interviewed at home every two years and we're about to go into the field with wave four this year. We identify the centres or schools that the children attend and mail out questionnaires. Roughly 7000 early childhood services have joined in and participated across the life of the study by filling in the questionnaires come to them, and returning them back for analysis.

Teachers in primary schools and childcare services are very involved in the LSAC study through the information that they provide.

The broad LSAC questions are looking at not just early childhood but family, school or community factors that impact on health, learning and social and emotional development.

Part of the family interview gives us very detailed information on what children's early childhood experiences are at the time when the interview is carried out. These were all done within a nine-month window of data collection in each of the main waves, when LSAC visits people in their homes. We get a little bit of information about the history but primarily it's their current information.

So for the K cohort, the four to five year old children who did that Who Am I? test that Amy Claessens analysed, we had very little information about their previous early childhood, but we had a lot of information about their current early childhood experiences.

Slide 10 - This next table summarises for just under 5000 children what their overall experience was. This is not just the main setting that they went to, but is based on information on up to three different early childhood current patterns of attendance at the time.

This table shows why I said most Australian four to five year olds seem to be attending a prior to school program. We know from this that, apart from the top two lines, all of those children were attending a centre based program, a preschool in a school – in a lot of the states, their preschools are attached to schools (WA, the ACT) - or a preschool in the school with extra child care. They might have been going to another centre as well or they may have also had home-based childcare from a relative. Those are children that have got not just their preschool program but also other additional care.

We also defined preschools that are not in a school. That's very typical in New South Wales or Victoria where preschools are typically not attached to schools. They're freestanding and run quite separately from the school system. Those children are either going to that sort of service alone or also attending other sorts of additional childcare programs.

Then there are the children in long day care settings, and long day care plus additional early childhood care.

So for children in the bottom six categories, we know they are all receiving some sort of centre based early childhood program before they start school. They're all four and a half to 5 years of age.

The top two groups need a little bit of thinking about. The very top one, that small group, five per cent of the overall sample, was not attending a centre-based program. Some of them were in family day care but most of them were not in any sort of formal program at all. They were in "exclusive parental care" which is the category we've used.

The next group though are the mystery group in a sense because they've already started school. We don't know, or we didn't know then at that time, where they were in the year before that - had they been to a preschool before they started school? We're trying to find that out but as far as I'm aware we haven't got it yet. (I'm not sure. Talk to the data people.)

Some of those children may have gone directly to school from home. Some of them may have gone via preschool. That anomaly is partly to do with the way that the states and territories organise the age of starting school and also to do with the LSAC study itself, which had quite a wide range of ages in the kindergarten cohort.

What I did was to look at that Who Am I? outcome. I asked "does early childhood have some impact on Who Am I?" scores at age four to five? As Amy Claessens did, and the other researchers who use the LSAC study, I looked at a whole lot of likely predictors and (slide 11) this is the set of the child and family characteristics that predicted the Who Am I? score at age four to five.

Not unexpectedly we see strong prediction from mothers' levels of education, the number of children in the families and child characteristics, gender, indigenous status, whether the child is growing up in a family speaking a language other than English. Children actually had higher scores on the Who Am I? if they were growing up in a family speaking a language other than English.

Next I looked at the early childhood education and care predictors. (slide 12) First of all, the children who had entered pre year one, not surprisingly, had significantly higher scores. In pre year one children tend to receive formal lessons in literacy and numeracy for at least half of their day.

That has to be kept in mind when we think about the other groups. But remember that 16 and a half per cent of the sample were in pre year one.

So for everybody else there are some interesting findings. I'll go through them with you here.

The first one is that, like EPPE, we found that there were benefits for the children who did attend a centre based program compared to the five per cent of children who did not attend a centre based program. The second point is that the benefits were seen for the children who attended these settings only. In other words they were at preschool only or long day care only. The children who also attended additional care were no different from the children who were at home in terms of their Who Am I? score.

That's telling us something about hours, weekly hours of childcare. It's not about the type. It's about the overall hours, the extra hours from going to other sorts of child settings as well as preschool or long day care.

Then taking account of all of the different types of care/early education, the scores were higher for children whose hours were mid range. That's part time hours, nine to 30 hours per week, which is very common in Australia. Most children are receiving part time experience in their early childhood programs. Those scores were higher, which again points to the lower scores for children over 30.

Then we looked at some of the quality indicators using information from the teachers who filled in a mail back questionnaire about their qualifications, their years of experience, their field of study, that is, did they have an early childhood qualification or a primary qualification or another area of study?

Then we asked about "approaches to learning" which is the time spent in the class on teacher directed whole group activities or teacher supported smaller group activities and child-initiated activities. Those categories were what we were able to gather in terms of the way that the classroom environment worked for the child.

We found marked differences between the effects of quality for children who had started school compared to the children who were in under five settings. Quality in a pre year one class contributed a pretty sizeable amount of variance compared to the overall variance. Five per cent was explained.

In the classrooms where the teacher organised for more time in teacher supported small group activities and less time in child initiated activities, children show higher scores on their Who Am I? test.

You may want to raise this later, but, as an early childhood lecturer, more time in child initiated activities is what we think teachers should be doing, that is, planning for children to initiate their own learning and follow their own interests.

However, in pre year one classes it worked in the opposite way. So I then asked my colleagues who work in schools, "What's going on here? How are school teachers interpreting this aspect of their program? What's likely to be going on in child-initiated activities?" They said "well it's that end of the day when children can just go off and choose things that they want to do. It's perhaps the unplanned time in pre year one classrooms. It's perhaps the free time, casual time where there's not a lot of concern given to organising or planning."

In contrast, free play is a very important part of planning in under-five settings.

So just coming back to the children in prior to school settings, the majority of the LSAC sample, there was very little effect of the measures of quality on Who Am I? scores, very minimal amount of variance was explained. One of the only things that showed a significant difference was the time spent in teacher directed large group activities. These are the activities where teachers are often planning very specific literacy, numeracy, activities, such as story time, news time.

The other finding is that across the different kinds of early childhood settings there was really very little difference in child outcomes between a preschool or a long day care centre program.

Slide 13. What are the unanswered questions? What is still very puzzling, having got this far with analysing LSAC Who Am I? scores for the four to five year olds in relation to their current childcare program,

is what predictors might be identified from their previous four years? Is it something to do with the patterns of childcare? for example, type of care, formal and informal care, amount of care per week? - We've had indications that the stability and changes of care over time are important. What about quality? LSAC was limited in that there was no capacity to gather observational information about quality. We relied on what teachers could tell us. Other studies are needed to look at how quality in Australian care compares to the US and UK studies by using a a standard measure that's used internationally, and then looking more specifically at whether there are particular aspects of quality that contribute to literacy and numeracy outcomes.

In order to address this I'm going to tell you about the Child Care Choices study which started a couple of years before LSAC and was able to inform some of the design work for LSAC.

This is a very different study though. (slide 14) It was specifically set up to look at the question of multiple and changeable childcare which was on the increase in the early 2000s. We wanted to investigate the use of multiple care, that is, care that was additional to the use of regulated long day care and family day care services. The study provides an interesting addition to what I've just presented in terms of the children attending preschool or long day care only, compared to those with additional care. There seemed to be some differences in their Who Am I? scores.

So as with many large studies of childcare, we used Bronfenbrenner's ecological model to inform the design of the study. (slide 15) What is different is that we looked at childcare in two ways. The yellow circle was what we were giving the most attention to. This was a study of children attending regulated childcare services, that's long day care and preschool and family day care.

In addition we knew that many of those children had multiple care, had extra care, so we separated that out in our overall design.

(slide 16) The background to the study is very similar to what I've already been speaking about, that adjustment and achievement in the first year of school, and where children are at at the age of four to five, is a key indicator of where they're going in the future. We wanted to look at the influences on these different areas of adjustment and achievement - the characteristics of the child, the family, but most particularly their non-parental childcare experience.

We collected information on the amount of care and the type of care - formal and informal care. Formal refers to regulated care settings and informal to care provided by babysitters or grandparents or drop-in care (in shopping centres) that's not subject to legislation.

We examined quality of care, multiple and changeable care, and a new area that we felt was very important as a measure of quality, the relationship that the child forms with his or her caregiver. This was a measure that was gathered by a questionnaire filled in by the child's primary carer in their child care setting.

Then we followed these children through to school and also looked at the impact of the classroom environment on these measures.

(slide 17) The research questions are looking at what in general predicts children's growth and development and wellbeing, adjustment and achievement.

In this presentation I'm only going to talk about the achievement outcomes, to follow on from the focus on literacy and numeracy measures that set children on track for their school success.

The way that we addressed the analyses of the data was to identify the best set of predictors, so rather than looking at everything we wanted to ask: "out of all of the potential predictors, what is the best set?" I'll probably skip some of the analysis explanations, but if anyone wants to know more about how we did that I can certainly share that with you.

Then we asked: "within that best set, what is the relative contribution of child, family, childcare and school predictors?"

Also of interest was the relative contributions of children's current preschool or childcare, preschool or childcare in the year before school, and their earlier childcare experiences.

In Slide 18 I'll show you the sample. We went to childcare centres and family day care services and we recruited all the families who were willing to participate whose children were under the age of three. Our sample ranged in age from four months, the youngest child in the study, to just over four years..

You can see that most of the children were in the infant/toddler or toddler age group, between ages one and three.

The people that provided information for us were the caregiver at home, almost always the mother but it could be another caregiver at home, the staff in the childcare services (because at this stage the children were all in childcare), including directors and the child's main teacher. As the children got older and moved into preschool, we then approached preschool teachers and preschool directors and kindergarten teachers once the children started school, (it's a New South Wales sample so kindergarten is the first year of school)

Recruitment was in regional and metropolitan centres in New South Wales so we had quite a varied sample of service types. We also recruited from a range of higher, lower and middle socioeconomic areas.

(slide 19) The process and procedure we used was an annual one. We visited the child in their childcare setting every year, once a year, and at that time the carers provided questionnaires and the research assistants conducted an observation of quality. If there was a child assessment to be done, the RA did that at the same time.

We used telephone interviews with parents through a CATI, computer assisted telephone interview, system.

(slide 20) These were the broad measures and those of you who are using the LSAC study or have looked at that will see that they're really very similar. We tried to cover all of the broad domains that are seen to be critically important when explaining how children are doing.

Many of these questionnaires were completed for us by parents but, where possible and relevant, we then also had the caregiver or the child's teacher filling in the questionnaire.

What's different from LSAC though is that we were able to go in and observe the centre using international measures to determine the quality of the program that children were attending. For the youngest ones, we used the Infant/Toddler Environment Rating Scale and as they moved into preschool or three year old rooms we used the Early Childhood Environment Rating Scale. These are the Harms and Clifford systems that have been used internationally.

Family day care also has an appropriate observation measure but it was harder to collect that information in family day care homes.

What was new was the measure of the child-carer relationship, the Student/Teacher Relationship Scale which can be used with carers of quite young children as well as classroom teachers.

(slide 21) The measures that we predicted to, and again I've put them all here but I'm just going to speak to the achievement measures today, were collected at school age and also the year before school, so four to five year olds and then also five to six year olds in their first year of school. The teacher and students themselves gave us information about how they felt about school and about their adjustment to school.

The achievement measures are different from LSAC but are picking up the same constructs. The Who Am I? is an early literacy and numeracy test. With Child Care Choices we used the Woodcock-Johnson battery of measures -  \ Applied Problems which picks up early mathematical thinking and literacy measures, Letter Word Identification and Spelling. These are measures that can be used as children get older so the Woodcock-Johnson tests provide a system where you can continue to collect information on children as they get older within the same method.

(slide 22) This is overview of the measures and the predictors and the outcomes. The red print are the key outcomes measures that we looked at at school age and also the year before school. Column 1 are groups of predictors from the early years. Because children came into the study at different ages, we created an average of these scoresup to three years of age.

Childcare or preschool history is summarised up toage three, and we also identify the specific amount of childcare preschool programs children are currently attending at the two time periods when we conducted the tests of achievement.

(slide 23) This is the analysis slide that I'll just speak to briefly. As is the case with many longitudinal studies, people move in and out of the study, or they don't fill in one questionnaire, but they fill in everything else, so many researchers are using multiple imputation techniques to maximise sample size. Not just a single imputation but multiple imputations so it's a "state of the art" analysis technique that my colleague, Alan Taylor, who is our statistician on the project, is very up to speed with.

The imputations were run as a two-step process: one to test each of the outcome measures by the 30 predictors to see the overall variance explained. Then two, because our aim was to identify the best set of predictors, was a reduction and then a multiple imputation test on the reduced model. This was a method that had been written about in the statistics literature and was a very acceptable way to tackle this challenge.

The first slide (24) shows the amount of variance that was explained for the five outcomes that we looked at; two in the year before school, the literacy and numeracy tests, and then in the first year of school the same literacy and numeracy tests and a measure of children's approach to learning that was rated by the teacher. This assessed level of motivation, independence, task management, the kind of skills that make a good learner in the classroom.

What I'll do is go through these slides. These are "best sets", showingwhat were the strongest predictors for each of these five measures.

(slide 25) For the numeracy test in the year before school, predictors were the age that the child was at the time of test, and their early abilities in numeracy. That was assessed using a similar test that was conducted when children were three year olds, so how they were doing then was the strongest predictor of how they were doing in the year before school in early mathematical thinking.

Like other studies we found that academic achievement is not just related to cognitive achievement, or cognitive skills and ability, but also related to social factors. Teacher or carer ratings of social behaviour also came in that best set of predictors, in a negative direction - children who showed fewer negative social behaviours were doing better in numeracy.

(slide 26) A year later the same two predictors (early ability in numeracy and social behaviour) are there, but age of the child is no longer significant in terms of the child's numeracy ability in the first year of school.

(slide 27) Looking at literacy we had more, quite a lot more, items staying in the set of predictors, and more domains identified as important predictors of literacy, measured by the letter word and the early spelling capabilities that children showed.

The family predictors were the number of children in the household. For families with larger numbers of children, the child did less well on the literacy scores. That's a finding that we certainly see in LSAC and other studies. So, larger numbers of children in the family is, in a way, a disadvantage for children's academic achievement. It's clearly an advantage in all sorts of other ways but in these particular measures (literacy and numeracy) it's a common finding that comes up.

Early development was the major area predicting children's literacy in the year before school. Again early ability in numeracy was a predictor. That was that measure that I mentioned before. Communicative ability, measured by the parent's rating of language and communication, was a positive influence. Negative social behaviours was a negative predictor of children's abilities in literacy.

Childcare history, for the first time, is included in this best set of predictors. The children who were attending fewer hours of formal care at the time of the test did better in their literacy achievement. Also children who had had a history of fewer multiple childcare arrangements.

(slide 28) The following year, as we saw with numeracy, the set of predictors changed a bit. Some predictors were no longer there. Age was no longer significant and childcare was no longer significant. Family factors were still important and the same behaviours - negative social behaviours and early ability in numeracy.

(slide 29) The last measure is the one that's to do with children's attitudes to learning, the measure of approach to learning in the first year of school. What we see here is the importance of family predictors and children's early abilities, their early developmental achievements in numeracy. Childcare history comes in as a predictor, in terms of less experience of formal childcare (centre based or family day care).

This is an important measure because many people would argue that the reason children do well in school is because of this sort of attitude to school. Thinking back to the beginning of the talk, and the quote from the Melbourne Declaration - "to build positive attitudes and competencies, the essential foundation for success in all learning areas."

So keeping this in mind, it could be that approach to learning is having an indirect effect via children's higher scores on literacy and numeracy.

(slide 30) This is an overall summary of the five different findings. Bold text shows the predictors that came up regularly across the two waves, the year before school and the first year of school. You can see pretty strong results showing that how children are doing at age three is predictive then of how they're doing at age four to five and how they're doing at age five to six.

(slide 31) This comes back to my question about trajectories starting earlier than we think. Some trajectories are hard to measure. You can't go down in age and do some of these tests with very young children. But the measure used in Child Care Choices is a very similar test to the Who Am I?. It's measuring the same constructs and we're seeing that how children achieve on that test at age three is setting them in train for how they are achieving two years later and then, from the LSAC data, we see that that trajectory continues through into primary school.

There's also an indication that history of childcare, less weekly hours of childcare, is a predictor. But you would have noticed (or perhaps not noticed because I haven't mentioned it) that nowhere in the results do we see quality of childcare coming up. It didn't come up as a predictor at all.

So what are some of the conclusions, questions, and where can I leave this and hand it over to you? The first sentence summarises that early cognitive abilities were consistently found to predict literacy and numeracy achievement in the year before school and the first year of school.

The childcare predictors, however, were those that were more aligned to families' choices. Families are the ones that choose to have or need to have longer hours of care. Families are the ones who have access to grandparents or friends or neighbours and so use multiple care for their child, for very good reasons. Families are the ones who can't access enough hours at their local preschool or their childcare so they add additional hours in other forms of care.

Those are the childcare predictors that are coming through in our analyses of this study. The quality predictors are not coming through. Why aren't we seeing an effect for the quality of early childcare attended by the child, particularly in this study where we assessed quality using a standardised international measure?

The next step, because every analysis leads to your next step, the next question, is do with the findings for early development. The early development measures are showing a strong trajectory. So the question is: Is child care quality absorbed by that result in some way? Is the quality effect hidden somehow by children's early developmental outcomes? Does quality predict the early development?

That's the next step that we'll be taking in the analyses of the Child Care Choices data. As we've done from the Who Am I? and LSAC, going back to age three, we now need to go back to the data collected prior to age three and see what we can find out about the question of child care quality.

(slide 32) I'll leave it there. I'll just show a couple of slides to acknowledge all the people that have worked on the Child Care Choices study, including Ben, who's sitting in front of me here, and then I said I would give you some of the websites so I'll just leave that one up for you if you want to copy any of that down. (slide 33)

The two studies that I've mentioned are available through online websites as also are all the COAG documents.

Thank you very much.

-- end --

IMPORTANT INFORMATION - PLEASE READ

The transcript is provided for information purposes only and is provided on the basis that all persons accessing the transcript undertake responsibility for assessing the relevance and accuracy of its content. Before using the material contained in the transcript, the permission of the relevant presenter should be obtained.

The Commonwealth of Australia, represented by the Australian Institute of Family Studies (AIFS), is not responsible for, and makes no representations in relation to, the accuracy of this transcript. AIFS does not accept any liability to any person for the content (or the use of such content) included in the transcript. The transcript may include or summarise views, standards or recommendations of third parties. The inclusion of such material is not an endorsement by AIFS of that material; nor does it indicate a commitment by AIFS to any particular course of action.

Top