Access to early childhood education in Australia

Research Report No. 24 – April 2013

You are in an archived section of the AIFS website. Archived publications may be of interest for historical reasons. Because of their age, they do not reflect current research data or AIFS' current research methodologies.

Appendix B. Measuring access to early childhood education services

This section provides an overview of how the various data sources have been used to analyses ECE in this report. Each of the following is described in turn: NSPCCC, AEDI, LSAC and CEaCS.

Within each section, in addition to a description of the data source, particular attention is given to the way in which ECE has been identified, and how children who were in the year before full-time schooling were identified, where applicable.

While the best solution to these and other measurement issues was sought, some problems remain, such that no dataset proved to be ideal in analysing ECE.

B.1 National Survey of Parents' Child Care Choices (NSPCCC)


This survey collected detailed information in 2009 about child care and ECE attendance of children aged 0-12 years.

The NSPCCC was designed to examine:

  • how parents make decisions around child care and preschool;
  • what key factors influence the use of child care and preschool;
  • how parents choose a child care service provider or preschool program; and
  • the impact of care choices of children reaching school age.

The survey was conducted by computer-assisted telephone interview, based on a random sample of Australian households, with sampling designed to obtain sufficient numbers of households with a child in the target age ranges (0-12 years of age). Households with children aged 3-5 years were over-sampled.

The sample frame for the NSPCCC survey was designed to record the child care status of a total sample size of 7,000 "reference children", with usually one, but sometimes two, reference children selected per household. In total, information was collected on 7,970 children from 5,625 households.

In selected households, the parent who was the main or joint decision-maker about child care or early education was identified and asked to provide details about the care arrangements for all children in the household aged up to 12 years. More details were then collected for the first child selected. Some details were collected for a second reference child, if applicable and if respondents agreed.

The sample was also selected with the aim of having minimum numbers of respondents in the various remoteness categories within each state.

Identifying children who were in the year before full-time schooling

Data were collected for a total of 7,970 children. Of these, 4,071 were aged 3-5 years, and were fairly equally spread across ages within this group. From these children, more detailed information was collected of those specifically identified as reference children. There were 3,865 3-5 year olds among the reference children.

To analyse ECE participation, details about the children's states of residence and birth dates were used to determine whether, at the time of the survey, they were in the year before full-time schooling. The eligibility information upon which this was based is shown in Table B1. The assumption was that children would go to school in the year that they became eligible to enter school. Children's actual attendance in ECE or school was also taken into account, such that if children were predicted to be in school (because their age made them eligible in that jurisdiction) but they were recorded as not yet being in school, then it was assumed they would start school the following year. Children were recorded as already being in school if they were reported to be in school and were aged 5 or over, even if their birth date and eligibility criterion led to their being coded as being in the year prior to full-time schooling.

Table B1: Assumptions regarding when starting school, NSPCCC
  Is in year before full-time schooling if: Number predicted to be in the year before full-time schooling
Age 4 years at survey Age 5 years at survey
NSW exact age at 31 July is 4 years to less than 5 years; or exact age at 31 July is 5 years, and not yet attending school 377 94
Vic. exact age at 30 April is 4 years to less than 5 years; or exact age at 30 April is 5 years, and not yet attending school 225 105
Qld exact age at 30 June is 4 years to less than 5 years; or exact age at 30 June is 5 years, and not yet attending school 229 47
SA age at time of survey is 4 years to less than 5 years; or age is 5 years, and not yet attending school 116 32
WA exact age at 30 June is 4 years to less than 5 years; or exact age at 30 June is 5 years, and not yet attending school 120 46
Tas. exact age at 31 January is 4 years to less than 5 years; or exact age at 31 January is 5 years, and not yet attending school 38 34
NT exact age at 30 June is 4 years to less than 5 years; or exact age at 30 June is 5 years, and not yet attending school 88 23
ACT exact age at 30 April is 4 years to less than 5 years; or exact age at 30 April is 5 years, and not yet attending school 40 23
Total   1,233 404

Note: A small number of children (N = 101) classified as 4 years old above were just less than 4 years old at the time of the survey, but were due to turn 4 years old before the cut-off date for that state/territory.

Given that some children are delayed in their entry to school, there are likely to be some children who are assessed in these data as being in the year before full-time schooling but who will actually be two years prior to starting schooling. As such, these data may overstate the percentage of children who are not in ECE in the year prior to full-time schooling.

Some children were excluded from the analyses, as it could not be determined whether or not they were already in school. This includes those who were classified as having no care/ECE (or school) arrangements, but when parents were asked why not, their reason was classified as "at school" (N = 266 children).

The NSPCCC subsample of reference children who were predicted to be in the year before full-time schooling was N = 1,637 children.

Classification of early childhood education

Parents were initially asked, for each child in the family, whether they were usually cared for by someone else. For children aged 4 and older, this was determined by the question: "I'd now like to ask you some questions about the care, preschool [or appropriate state nomenclature], kindergarten [or appropriate state nomenclature], out-of-school hours or school arrangements for your child/children. Apart from you [you and current partner] does anyone else look after or supervise [child] on a regular basis? This includes both teachers and care workers." If the response was "no", this was the end of the care/ECE questions and details were asked of the next child, if applicable. While this question quite specifically referred to teachers as well as care workers, there may have been some children coded here to "no" who were actually in some ECE program. The NSPCCC methodological report (Social Research Centre, 2010) states:

As expected, there was some respondent confusion around the relationship between childcare, pre school and school. Even with the additional scripting to focus parents on care provided by teachers, many respondents did not relate school or pre school to child care and, as such, were asked additional questions that they did not feel were directly relevant to their situation. (p. 21)

Parents who answered "yes" to this screening question, for each child, were then asked, "Is care provided by a child care centre, a carer or a teacher or does someone else look after the child?". If the response was "yes", parents were asked about the following types of formal care: family day care, home care, LDC at a child care centre, occasional care, and after-school-hours care. When children were said to attend LDC, parents were further asked if the child received a preschool education program while there. Parents who did not know if their child attended a preschool program in LDC (5% of children in LDC in the sample analysed) were included with LDC without preschool. Other details of informal care use were also collected.

Parents of children aged 3-5 years, were then asked if the child attended preschool or received a preschool education program. Those who already said their child attended a preschool education program in LDC were asked if this child attended preschool somewhere else.

From this information, children were classified as shown in Table B2. Children classified as having no preschool/LDC includes children only in parental care, family day care, receiving home care or other informal care. Note that in this report, for simplicity of presentation, we usually present LDC at the aggregate level only, omitting the detail of whether it was with or without a preschool program.

Table B2: Classification of early childhood education using NSPCCC reference children
  N Unweighted % Weighted %
Preschool only 757 46.2 42.3
Preschool and LDC 159 9.7 8.1
LDC 402 24.6 31.7
without a preschool program 93 5.7 7.3
with a preschool program 309 18.9 24.4
Not in ECE 319 19.5 17.9
Total 1,637 100.0 100.0

Interviewers were given instructions about the typical characteristics of LDC and preschool/kindergarten. They were instructed about the different nomenclature for preschool and for the pre-Year 1 level of schooling. The state-appropriate nomenclature was used in the questionnaire.

Children were excluded from the analyses if the type of early education/care could not be determined.

These details were collected of all children in the family. For the more detailed questions in NSPCCC, one reference child (or sometimes two) was selected from each family, with various rules given to interviewers in how these reference children were to be selected (e.g., "priority 1" children were those aged 4-5 years, to maximise the number of reference children in these years of early education). More detailed information about care/ECE was collected for these reference children.

This report uses information for these reference children, rather than all children in the family.

Summary of key data limitations and problems

  • Some children may have been coded as having no care (or early education), when in fact they were in some form of care or early education.
  • Some 4-year-old children classified as being in the year before full-time schooling may in fact be two years before full-time schooling, and so the percentage in no ECE may overestimate the percentage among children who are in the year before full-time schooling.
  • While the sample size is reasonable, even at the state/territory level, the sample sizes become very small when wishing to analyse specific patterns (or non-use) of ECE by demographic groups.

B.2 Growing up in Australia: The Longitudinal Study of Australian Children (LSAC)


Growing Up in Australia: The Longitudinal Study of Australian Children is being conducted in a partnership between FaHCSIA, AIFS and the ABS. The study aims to examine the impact of Australia's unique social, economic and cultural environment on children growing up in today's world.

The study follows two cohorts of children who were selected from across Australia. Children in the B cohort ("babies" at Wave 1) were born between March 2003 and February 2004, and children in the K cohort ("kindergarten" at Wave 1) were born between March 1999 and February 2000. The study is representative of Australian children in these birth cohorts.

The sampling frame for LSAC was taken from the Medicare database. Initially, a sample of postal areas was chosen, and children in the birth cohorts in those regions were selected. Postal areas from across Australia were selected, with the exception of remote parts of Australia.

To date, data from four main waves of the survey are available, collected in 2004, 2006, 2008 and 2010.

Much of the information in LSAC is collected from the child's primary carer, who is usually the mother.

This report uses information collected about the B cohort, when the children were aged 4-5 years, in 2008. This is Wave 3 of LSAC. The total sample size at that time was 4,386 children. Note that there is some bias in this sample, with the Wave 3 sample comprising around 86% of the original Wave 1 sample. Sample weights adjust for the probability of selection into the LSAC sample, as well as attrition, and have been used in these analyses.

While the K cohort at Wave 1 could also be used to provide information about care and ECE arrangements of 4-5 year olds, these data were collected in 2004, and so it was considered preferable to use the B cohort data that provided information for this age group at a more recent time. Questions about ECE and care varied between the two cohorts, as the B cohort questions were improved to address some problems that were identified with the Wave 1 collection (LSAC Project Operations Team, 2009).

Identifying children who were in the year before full-time schooling

Children in the B cohort, at the third wave, were aged 4-5 years, and were therefore at the age of preschool or school participation. For these analyses, children were excluded if they were already in pre-Year 1 in school. This left a possible sample of 3,005 children. All these children were due to start full-time schooling the next year, according to parents' reports of the program that their children were expected to attend the following year.

Classification of early childhood education

Considerable effort was made in this study to design questions that captured the various ways in which care and ECE can be provided at this age, and to cater for the state differences in nomenclature. Interviewers were provided with supplementary information about the state variation in names for preschool and for the first year of full-time schooling (LSAC Project Operations Team, 2009).

The primary carer was initially asked "Does the Study Child currently go to a school, kindergarten, pre-school or a long day care centre?". If yes, they were asked if this was in a school. (If the children attended more than one, the parents were asked to report about the one in which the children spent the most hours per week.)

Parents were asked what program this was, including Year or Grade 1, pre-Year 1 program, preschool/kindergarten, LDC centre/early learning program or other. For children not in school, parents were similarly asked about the program the child attended, and they were given options of preschool/kindergarten-only centre; preschool/kindergarten in a LDC centre; mobile preschool; LDC; and other.

Quite a lot of the LSAC parents chose "LDC" rather than "preschool program in LDC". We are reluctant to interpret this as meaning those who chose "LDC" did not have a preschool program provided. It may simply be that parents see this predominantly as LDC. No particular probing followed to check whether a preschool program was offered. We have therefore classifed these respondents just as "LDC".

Detailed information about this main care/ECE program was collected in LSAC, and these details have been used when presenting such information in this report.

Later, parents were asked about other care arrangement of the child. This information was used in conjunction with the above information to determine whether children attended a mix of care arrangements, to identify children who attended preschool as well as LDC. Other details of these secondary arrangement (e.g., hours in care) have not been used in this report.

The information was used to classify children as shown in Table B3. Note that in this report, for simplicity of presentation, we usually present LDC at the aggregate level only, omitting the detail of whether or not this was with a preschool program.

Table B3: Classification of early childhood education using LSAC
  N Unweighted % Weighted %
Preschool only 1,684 56.0 54.8
Preschool and LDC 321 10.7 9.6
LDC 821 27.3 28.4
LDC (with preschool program not selected) 346 11.5 11.6
LDC with a preschool program 475 15.8 16.8
Not in ECE 179 6.0 7.1
Total 3,005 100.0 100.0

Summary of key data limitations and problems

  • LSAC does not cover remote parts of Australia. State estimates can be produced, although sample sizes are quite small in the smaller states. State estimates will be differentially affected by the exclusion of remote areas.
  • The bias in the sample, largely due to attrition at Wave 3, means that the most disadvantaged Australian children may not be fully represented in the study.

B.3 Australian Early Development Index (AEDI)


The AEDI is being conducted by the Centre for Community Child Health (at The Royal Children's Hospital, Melbourne) in partnership with the Telethon Institute for Child Health Research, Perth.

The AEDI measures how young children are developing in defined Australian "communities". It provides a population measure of children's development at the time they start primary school. The primary aim of the AEDI is to provide data to help communities in the development and reorientation of services and systems to enable them to improve the health and wellbeing of young children.

Data are to be collected every three years, with the first data collection taking place in 2009. The AEDI population comprises all children in the first year of full-time primary schooling within a community or a geographic area.

The AEDI is based on the scores from a teacher-completed checklist, with over 100 questions covering five developmental domains: physical health and wellbeing, social competence, emotional maturity, language and cognitive skills, and communication skills and general knowledge. The analyses in this paper are based on the full AEDI unit record file.

Identifying in-scope sample

To focus the analyses on children who were eligible for ECE in 2008 (the year prior to the 2009 collection, of children in the first year of full-time schooling), children were excluded if (a) they were included in the 2010 AEDI supplementary collection; or (b) they were recorded as having repeated the first year of full-time schooling. This left a total sample of 256,817 children, out of the total 268,726 children on the unit record file.

Notionally, all of these children should have been eligible for early education. Exceptions may exist, for example, for children who have very recently migrated to Australia; however, these data were analysed on the assumption that all AEDI respondents will have been eligible for early childhood education.

Classification of early childhood education

For each child, information about program type in the year before entering school is requested in the AEDI collection. For teachers completing the checklist, this information will have been sourced from school enrolment forms, and so relies upon parents having completed this information with sufficient detail to enable the data to be provided by the teacher.

The first question asked is, "In the year before entering school has the child been in non-parental care on a regular basis and/or attended any other educational programs?". The overall non-response ("don't know", or missing) to this item was 7.9% (among those in the in-scope sample for these analyses). Non-respondents were excluded from the analyses, reducing the in-scope sample to 236,284.

The items available for analyses then classify children into "yes", "no" or "don't know" for the types of care or ECE they received prior to full-time schooling. There were some additional non-responses here. A total of 50,664 children were reported to have attended a day care centre with a preschool/kindergarten program, and 15,730 attended a day care centre without a preschool/kindergarten program, but 29,458 children were reported to have attended a day care centre where it was not known if this included a preschool/kindergarten program. This meant the distinction between with/without a preschool/kindergarten program was somewhat unreliable. As a result, this distinction is not used in these analyses. A very small number of records were excluded at this point, as missing data on these items meant the type of ECE prior to starting full-time school could not be derived.

Note that there was somewhat higher reporting of having no prior ECE than might have been expected in some states. One possibility is that children who received ECE at the school were coded as not being enrolled elsewhere prior to school commencement, and so estimates may undercount ECE participation. This information was used to classify children as shown in Table B4.

Table B4: Classification of early childhood education using AEDI
  N Unweighted %
Preschool only 128,545 54.4
Preschool and LDC 24,055 10.2
LDC only 58,252 24.7
Not in ECE 25,399 10.8
Total 236,251 100.0

Note: As the AEDI covered the vast majority of children in Year 1 of school, no weights were used in analyses.

Summary of key data limitations and problems

The key limitation of these data relates to the uncertainty of the data quality. Data quality relies on:

  • parents providing schools with accurate information on ECE participation;
  • this information allowing identification of preschool versus child care;
  • how ECE that children attended at the school prior to pre-Year 1 school was captured; and
  • teachers referring to this information when completing the AEDI checklist.

Also, the AEDI data are limited in respect to family information, such that ECE participation cannot be related to parental employment, education, income or relationship status.

It must be noted, however, that the very obvious strength of these data lie in the size of the dataset and coverage across Australia, which allow analyses of the more vulnerable groups of children, as well as state-level analyses.

B.4 The Childhood Education and Care Survey (CEaCS), ABS

The ABS has conducted a survey of child care use approximately every three years since 1969. This was previously (up until 2005) known as the Child Care Survey. Its name was changed in 2008 to reflect its increased focus on education.

The CEaCS is conducted in conjunction with the monthly labour force survey, which is a multistage area sample of private and non-private dwellings across Australia. Information is collected by interview with one person in the family within selected households, with information most often reported by mothers. Data collected relate to child care use in the week prior to the interview. This is timed to be outside of school holidays.

Within respondent households, child care details are collected for children aged under 12 years old. In families with more than two children in a family, two children are randomly selected and detailed child care items are collected for these children. (The application of weights ensures their representativeness of all children.)

The survey data were used in this report to analyse parental decision-making and preferences concerning children's non-participation in ECE, or participation in different types of ECE.

While these data could also have been analysed to examine ECE participation rates, we could not accurately identify children who were in the year prior to full-time school, so the analyses would not have been consistent with analyses of other datasets. We were also aware that these data were being analysed in respect to ECE participation in another report under preparation.

The information on school-aged children's prior participation in ECE could have been used, but we felt that analyses of the three other datasets provided us with sufficient information to address the research questions. These data were not as current as the other data included, which would have been especially problematic considering the changes in ECE that have occurred in Queensland in recent years.