This document was added or updated on 02/10/2015.

FACTORS INFLUENCING EARLY CHILDHOOD DEVELOPMENT IN TASMANIA

INTRODUCTION

Early childhood development is of central importance to the wellbeing of Australia's children and the future productivity of the nation.¹ Evidence tells us that a child's early development can shape their later life outcomes.² Studies have shown that early childhood education and socioeconomic background play an important role in a child's development and successful transition into school.³ Following on from the previous ABS study focusing on Queensland,⁴ this article presents analysis on the extent to which preschool participation and parental, family and household characteristics affect child development in Tasmania.

The Australian Early Development Census (AEDC, formerly the Australian Early Development Index) is a population measure of children's physical, social, emotional, communication and cognitive development as they enter school. While the AEDC provides some demographic data, it contains limited parental, family, household or early childhood education information.

The information in this article has been compiled by linking data from the ABS Census of Population and Housing (Census) with AEDC and National Early Childhood Education and Care Collection (NECECC) data for Tasmania. The integrated dataset supports analysis on the extent to which preschool participation (as measured by 2011 NECECC data) and personal and socioeconomic characteristics (such as family composition, parental education and household income from the 2011 ABS Census) affect child development (measured by the 2012 AEDC).

A logistic regression model was built to determine whether selected characteristics had an effect on the likelihood of a child being developmentally vulnerable. The regression analysis allows the influence of each factor on a child's development to be isolated. Some results of this analysis are presented throughout this article and the full results can be found in the Appendix: Factors influencing early childhood development in Tasmania - Results of regression analysis.

This integrated approach leverages more information from the combined dataset than is available from the individual datasets. Importantly, it enables better understanding of child development across population groups, which is important for improving policy development, service delivery and evaluation. This article demonstrates how the integrated dataset can enhance the evidence base for social, economic and education policy in Australia. It also provides a basis for future research.


DATA IN THIS ARTICLE

The AEDC is completed by teachers for children in their first year of full-time schooling (known as a Preparatory year or 'Prep') in Tasmania. The AEDC measures five areas or 'domains' of early childhood development:

• Physical health and wellbeing - whether children are healthy, independent, and physically ready for the school day, as well as their gross and fine motor skills
• Social competence - children's overall social competence as well as how they play, share and get along with other children
• Emotional maturity - whether children are able to concentrate during the school day, help others, are patient and not aggressive or angry
• Language and cognitive skills - mainly based on those skills necessary for school, including literacy, numeracy and memory
• Communication and general knowledge - whether children can communicate easily and effectively, and have adequate general knowledge.⁵

Domain scores are calculated based on teacher responses to a range of developmental and observational questions on each child. Through this process some children are classified as 'developmentally vulnerable'.⁶ Whilst the majority of children in Tasmania were doing well on each of the five developmental domains in 2012, 22% of children in Prep were developmentally vulnerable on one or more domains, and 10% were developmentally vulnerable on two or more domains.

While the AEDC provides information on children in their first year of formal schooling, the NECECC was used in this study to provide data on children enrolled in preschool programs (commonly referred to as 'Kindergarten' in Tasmania) in the year prior to formal schooling. For the purpose of the NECECC, a preschool program is defined as a structured, play based learning program, delivered by a degree qualified teacher, aimed primarily at children in the year or two before they commence full-time schooling.⁷ Participation in preschool is not compulsory.

The ABS Census of Population and Housing is conducted every five years and aims to accurately measure the number and key characteristics of people who are in Australia on Census Night, and of the dwellings in which they live. Due to ABS Census data being collected in August 2011 and AEDC data being collected between May and July 2012, there is a possibility that some children's family and household characteristics may have changed between the time of the ABS Census and the AEDC.

The data used in this article is based on linked AEDC and ABS Census records. The NECECC data was included in the dataset where a corresponding ABS Census record was able to be linked. Records were linked across these datasets by finding exact matches on combinations of common variables, such as date of birth, sex, and small area geography codes. Name and address information were not available for linkage. The data has been weighted to ensure it is representative of the full Tasmanian AEDC population. There may be differences between figures in this article and those in publications that use the individual datasets or other data sources.

For more detailed information about data sources, definitions and linkage methodologies, see the Explanatory Notes tab.


PRESCHOOL PARTICIPATION

Effect of time at preschool varies

Children enrolled in preschool for 20 hours or more per week in the year before school (2011) had the lowest proportion of developmental vulnerability on two or more domains (4%), followed by those enrolled for 11 to 14 hours (8%). Children enrolled in preschool for between 1 and 10 hours per week had the highest proportion of developmental vulnerability on two or more domains (14%).


Looking at each domain individually, preschool participation appears to have different effects. The logistic regression model results show that while children enrolled for 20 hours or more per week were significantly more likely than those enrolled for 11 to 14 hours to be developmentally vulnerable on the Physical health and wellbeing domain, they were significantly less likely to be vulnerable in the Language and cognition domain. In fact, none of the 234 children on the integrated dataset who were enrolled in preschool for 20 hours or more per week were classified as developmentally vulnerable on the Language and cognition domain. Children enrolled for 1 to 10 hours of preschool per week were around twice as likely to be developmentally vulnerable in the Social competence and Emotional maturity domains than children enrolled for 11 to 14 hours. Children enrolled for 15 to 19 hours of preschool per week were significantly more likely to be developmentally vulnerable than those children enrolled for 11 to 14 hours in the Physical health and wellbeing, Social competence and Language and cognition domains.

Parents who don't work use more preschool

Children from couple families where all parents were not in the labour force had the largest proportion enrolled for 15 to 19 hours of preschool per week (53%). In contrast, children living in couple families with both parents employed or living with an employed lone parent ('all employed', regardless of full-time or part-time status) had the largest proportion enrolled in preschool for 11 to 14 hours per week (42%).


Tasmania commenced implementing the Universal Access to Early Childhood Education (known as '15 hours of kindergarten') in schools in 2010, with all schools offering 15 hours of preschool by 2014. By 2012, 90% of children were enrolled in a kindergarten program offered for 15 hours or more per week.⁸


PERSONAL CHARACTERISTICS AND CIRCUMSTANCES

Regularity of reading to children affects their development

The AEDC collects information on a number of general characteristics and behaviours which do not directly contribute to a child being classified as developmentally vulnerable. The integrated dataset allows analysis of the extent to which responses to individual AEDC questions relate to child development when holding factors obtained from the ABS Census and the NECECC constant. One such question asks teachers to assess the regularity with which a child is read to, or encouraged in their reading, at home. This may also take into account the child talking about the books they have read at home and going to the library regularly, and parent(s)/caregiver(s) reporting to teachers that they hear their child reading regularly at home.

The regression modelling results show that, when holding other factors constant, how regularly a child was read to had a strong and consistently significant effect on developmental vulnerability. Across the domains, children who were assessed by teachers as not being read to regularly at home were around three times more likely to be developmentally vulnerable compared with children who were read to somewhat regularly at home. In addition, children who were assessed as being read to very regularly were significantly less likely to be developmentally vulnerable than those who were somewhat regularly read to, for each of the domains.

Parental engagement has a strong relationship with child development

The AEDC collects information on parental engagement with a child's school in support of their child's learning. This may include the parent(s)/caregiver(s) speaking to the teacher about their child’s learning or concerns they may have about their child, and attending parent/teacher information nights or interviews at the school. The regression analysis shows that parental engagement with a child's school had a strong and consistently significant relationship with a child's developmental vulnerability, with higher levels of parental engagement related to better outcomes. Children whose parents were very engaged with the school were 75% to 85% less likely to be developmentally vulnerable across the domains than those whose parents were only somewhat engaged. In contrast, children whose parents were not engaged with the school were two to three times more likely to be developmentally vulnerable than those whose parents were somewhat engaged.

Boys more likely to be developmentally vulnerable

Consistent with national AEDC results,⁹ boys were significantly more likely than girls to be developmentally vulnerable across all five domains. This was particularly the case with the Emotional maturity domain, where the regression analysis shows that boys were over four times more likely to be developmentally vulnerable than girls.

Most Aboriginal and Torres Strait Islander children not developmentally vulnerable

Consistent with national AEDC results,¹⁰ the majority of Aboriginal and Torres Strait Islander children in Tasmania were not developmentally vulnerable on any of the domains. Despite this, nearly double the proportion of Aboriginal and Torres Strait Islander children were developmentally vulnerable on two or more domains (17%) compared with non-Indigenous children (9%). However, when other factors such as remoteness and parental education were held constant, Aboriginal and Torres Strait Islander children were significantly more likely than non-Indigenous children to be developmentally vulnerable on Language and cognition only. There were no significant differences in developmental vulnerability between these two groups on any other domains.

Natural or adopted children don't always do best

Census data gives us the ability to be able to distinguish between different kinds of children and the family construct in which they live. The integrated dataset enabled the comparison of results for children who were the natural or adopted child of the parent(s) they were living with, and children who were living in families where they were the step-child of either parent, foster child, grand-child, otherwise related child or unrelated child. The regression results show that natural or adopted children were significantly less likely than those living with at least one adult who was not their natural or adopted parent to be developmentally vulnerable on the Emotional maturity domain. Interestingly, when compared with those children living with at least one adult who was not their natural or adopted parent, natural or adopted children were over twice as likely to be developmentally vulnerable on the Communication and general knowledge domain.

Developmental vulnerability differs for younger school starters

In Tasmania, the compulsory school starting age is five years, meaning that a child who is at least five years old as at January 1 must be enrolled in full-time Preparatory at a school or provided with home education, unless exemptions apply.¹¹ When other factors were held constant, children who started school aged younger than five and a half years were significantly more likely to be vulnerable on the Physical health and wellbeing domain (which includes items on toileting habits, physical energy throughout the day and proficiency at holding a crayon) than those who started school aged five and a half to less than six years. In contrast, they were significantly less likely to be vulnerable on the Language and cognition domain, which assesses a child's ability to read complex words, count to twenty and understand simple time concepts. However, it is important to note that there can be individual variability in the specific timing at which children achieve developmental milestones, and any delays in school starting age do not necessarily equate to having later academic difficulties.¹²


FAMILY CHARACTERISTICS

Children with highly educated parents less likely to be developmentally vulnerable

The ABS Census provides detailed information about the level (and field) of parents' education that is not currently available on the AEDC. Where a child lives with both parents, this article presents details of the parent with the higher education level.

Generally, the higher the parents' education levels, the less likely their children were to be developmentally vulnerable. However, the regression analysis suggests that when all other variables are held constant, this effect was generally only significant when comparing the outermost groups, such as those with the highest parental education at the Postgraduate or Graduate level compared with Year 11 or below, and only on some domains.

Children of employed parents less likely to be developmentally vulnerable

Children living in couple families with both parents employed or living with an employed lone parent (referred to as 'all employed', regardless of full-time or part-time status) had the lowest proportion of developmental vulnerability across all domains.

However, the regression modelling results show that the effect of parental employment status was only significant on the Language and cognition domain, where children in couple families with one employed parent were more likely to be vulnerable than children with all parents employed, and the Communication and general knowledge domain, where children in couple families with one employed parent or in families with all parents not in the labour force were more likely to be vulnerable than children with all parents employed.

Vulnerability varies across parental occupations

While children of employed parents were less likely to be classified as developmentally vulnerable, their rates of vulnerability differed considerably across different parental occupations. When looking at the top 30 most common occupations for fathers of children in Tasmania, the lowest rates of developmental vulnerability on two or more domains were for children of Construction, Distribution and Production Managers or Defence Force Members, Fire Fighters and Police (both 2%), while the highest rate of developmental vulnerability on two or more domains was for children of Farm, Forestry and Garden Workers (22%).

When considering the top 30 most common occupations for mothers, there were no children classified as developmentally vulnerable on two or more domains if their mother had one of the following occupations: Human Resource and Training Professionals, Information and Organisation Professionals, Natural and Physical Science Professionals, Receptionists, and Tertiary Education Teachers. The highest rate of developmental vulnerability on two or more domains was for children of Personal Carers and Assistants (17%). It should be noted that parental occupation was not included in the regression model and it has therefore not been determined if the relationship between parental occupation and child development is significant.

Family size plays a role in child development

Across most domains, children living in families with two or three children had the lowest proportions of developmental vulnerability, while children living in families with one child or five or more children had the highest proportions of developmental vulnerability.


After controlling for other factors, such as income and family type, having more siblings generally had a significant positive effect on developmental vulnerability in several domains. These effects were most consistent in the Social competence and Emotional maturity domains, where children living in families with three or more children were significantly less likely to be developmentally vulnerable than those from families with two children.

Birthplace of parents related to physical health and wellbeing

The regression modelling results show that children with both parents born overseas were 65% less likely to be developmentally vulnerable on the Physical health and wellbeing domain than those with both parents born in Australia. Parental birthplace did not have a significant impact on children's developmental vulnerability for any of the other domains.

Developmental vulnerability differs across family types

Children of married couples had a lower proportion of developmental vulnerability on two or more domains (6%) than those of de facto couples (13%) or lone parents (20%), however this difference was more prevalent for girls. For example, the rate of developmental vulnerability on two or more domains for girls living in married couple families was over five times lower than for those living in one parent families, whereas the rate for boys was three times lower. Interestingly, when holding other factors such as sex constant, family type did not have a significant impact on developmental vulnerability for any of the five domains.


Mother's age has little impact on child's developmental vulnerability

Generally, across all domains, children of older mothers had lower rates of developmental vulnerability than children of younger mothers. However, the regression results show that when holding other factors constant, mother's age at the time of their child's birth did not have a significant effect on developmental vulnerability for any of the five domains.

HOUSEHOLD CHARACTERISTICS

Lower rates of developmental vulnerability for children in higher income families

In general, higher levels of household income were associated with lower proportions of children being developmentally vulnerable on two or more domains. Children living in a household with an income of $2,500 or more per week had a rate of developmental vulnerability on two or more domains approximately three times lower than children living in a household with a weekly income of less than $600.


However, the regression analysis shows that the impact of household income is generally not significant when holding other factors constant. The only significant results were in the Emotional maturity domain, where children living in households with a weekly income between $1,500 and $2,499 were more likely to be developmentally vulnerable than those in a household with an income of $1,000 to $1,499 per week.

Vulnerability varies within socioeconomic areas

Whilst developmental vulnerability may be analysed based on the socioeconomic status of the area in which the child lives (as measured by Socio-Economic Indexes for Areas, or SEIFA), a benefit of the integrated dataset is that socioeconomic measures can be explored at a household or family level within each SEIFA quintile. For example, in the most disadvantaged areas (Quintile 1) the proportion of children who were developmentally vulnerable on two or more domains varied from 23% of children in households with an income of less than $1,000 per week, to 12% of children in households with an income of between $1,000 and $1,999 a week. Likewise, in the most advantaged areas of Tasmania (Quintile 5) the proportion of children developmentally vulnerable on two or more domains varied from 9% of children in households with an income of less than $1,000 per week, to just 3% of children in households with a weekly income of $2,000 or more. The interaction between SEIFA and household income was not included in the regression model.

(a) Based on the 2011 Socio-Economic Indexes for Areas (SEIFA) Index of Relative Socio-economic Advantage and Disadvantage (IRSAD). (b) Includes negative or nil income.

Parental employment and education were beneficial in all socioeconomic areas

Across all SEIFA quintiles, having at least one parent employed was related to lower rates of developmental vulnerability. Children living in the most advantaged areas (Quintile 5) had the lowest rates of vulnerability even when no parents were employed, while those children living in the most disadvantaged areas (Quintile 1) had the highest rates of vulnerability on two or more domains even with all parents employed.


Children living with a parent who had a Bachelor Degree or higher generally had lower rates of developmental vulnerability, regardless of which SEIFA quintile they lived in.


Children in Outer Regional Tasmania less vulnerable

Throughout much of Australia, outcomes for people living in Outer Regional, Remote or Very Remote areas are poorer than for those living in Major Cities or Inner Regional areas.¹³ However, this was not the case in Tasmania, where it appears that children living in Outer Regional areas had a lower proportion of developmental vulnerability across some domains. Reasons for this may include the relatively small geographic area of Tasmania, which means that Outer Regional areas are not so far from Inner Regional areas, resulting in more of a community environment. There is also evidence of "tree changers",¹⁴ which may mean the characteristics of people living in the Outer Regional areas of Tasmania are different to those living in the Outer Regional areas of other states or territories.


The regression analysis shows that children living in Outer Regional areas of Tasmania were significantly less likely than those living in Inner Regional areas to be developmentally vulnerable on the Physical health and wellbeing, Social competence and Emotional maturity domains. The apparent differences between children living in Remote or Very Remote Tasmania and those living in Inner Regional areas were found to be not significant when other variables were held constant.

Overall in 2012, 10% of children in Tasmania were developmentally vulnerable on two or more domains. However, children in some areas of Tasmania fared better than others, with the Statistical Area 3 of 'Meander Valley - West Tamar' having the lowest proportion of children developmentally vulnerable on two or more domains (4%). At the other end of the spectrum, 23% of children living in the Statistical Area 3 of 'Brighton' were developmentally vulnerable on two or more domains.

DATA COMPARISONS In addition to enhancing the information available from individual datasets, integrated data is also able to provide comparisons between variables that exist on multiple datasets. This enables the quality and consistency of reporting on similar variables to be checked. The integrated Tasmanian Education and ABS Census dataset allowed an assessment of the quality and consistency of Preschool attendance/enrolment and Indigenous status across the three datasets, that is, the Australian Early Development Census (AEDC), National Early Childhood Education and Care Collection (NECECC) and ABS Census. The quality of these items are of particular interest to data custodians, including for measurement of targets related to national initiatives, such as the National Agreements and National Partnerships, and national reporting. Preschool reporting There was very high consistency of reporting of preschool attendance between the three datasets for Tasmania. Almost 95% of the children identified in the AEDC as having attended preschool were also identified in the ABS Census as attending either a preschool or Infants/Primary school in 2011. In addition, of those identified in the AEDC as having attended preschool, 90% were also identified in the NECECC as having been enrolled in preschool. Of the remainder, 0.2% were identified in the NECECC as not being enrolled in preschool while 9.8% were not successfully linked to an NECECC record. It should be noted that Tasmania had very high proportions of children attending government funded preschools in 201115 and the consistency of reporting seen in these three datasets in Tasmania may not hold for other states and territories. Overall, between 90% (in the NECECC) and 95% (in the AEDC) of children in Tasmania were identified as attending preschool. Indigenous status reporting There is generally good consistency between reporting of Indigenous status in the AEDC and ABS Census in Tasmania. Approximately 87% of children identified in the AEDC as being of Aboriginal and/or Torres Strait Islander origin were also identified as being an Aboriginal and Torres Strait Islander child in the ABS Census in 2011. Reporting was slightly less consistent between AEDC and NECECC, with 84% of children identified as being of Aboriginal and/or Torres Strait Islander origin in the AEDC, also identified in the NECECC. Of the 518 Aboriginal and Torres Strait Islander children identified in the ABS Census data within the integrated dataset, which had the highest rate of identification, 71% were identified as being of Aboriginal and/or Torres Strait Islander origin on all three datasets.

CONCLUSION

Using Tasmanian administrative data from the Australian Early Development Census (AEDC) and National Early Childhood Education and Care Collection (NECECC), in conjunction with ABS Census data, this article provides evidence of the socioeconomic and contextual factors that influence developmental vulnerability. When other factors were held constant, the regularity with which a child was read to (or encouraged in their reading) at home, parental engagement with a child's school, and a child's sex all had strong and consistent relationships with their developmental vulnerability across all AEDC domains.

Other factors which had a significant effect on a child's developmental vulnerability in their first year of schooling across some domains included the number of hours a child was enrolled in preschool in the previous year, parental education, the number of siblings a child had, the age at which a child started school, child type, and the socioeconomic status and remoteness of the area in which they lived. Factors which only had a slight effect were parental labour force status, household income, parental birthplace, and Indigenous status. Factors which did not have a significant effect on any domain were family type, mother's age at the time of the child's birth, tenure type, need for extra bedrooms, whether the child was cared for by a relative in the year before starting school, and type of internet access.


LOOKING AHEAD

There is extensive scope to enhance the evidence base for social, economic and educational policy by maximising the use of existing administrative data in conjunction with collections such as the Australian Early Development Census and the ABS Census of Population and Housing. Further analysis of the results presented in this article could be beneficial in better understanding why certain factors appear to influence developmental vulnerability, at times in contradictory ways. Expanding this analysis to cover national data rather than a specific state would also be beneficial to determine if factors associated with developmental vulnerability vary across Australia, as well as to improve the ability to report on small population groups.

Recent data integration work performed by the ABS has linked data on student literacy and numeracy achievement as measured in NAPLAN testing to the Census of Population and Housing.¹⁶ Future research could look at integrating the AEDC and Census data with data from NAPLAN to provide information on the relationship between children's early development and later achievement, and how these are moderated by various parental and socioeconomic contextual factors. Particular areas of interest might include how early childhood development and educational outcomes can be optimised for children from disadvantaged backgrounds.


ENDNOTES

1. AEDC, How the AEDC assists policy reform, Accessed 15 July 2015 2. AEDC, The importance of early childhood development, Accessed 15 July 2015; and Brinkman, S et al. 2014, The predictive validity of the AEDC: Predicting later cognitive and behavioural outcomes, Accessed 20 August 2015 3. Goldfield, S et al. 2014, Early childhood education and care and the transition to school; AEDC, The impact of socio-economics and school readiness for life course educational trajectories, Accessed 15 July 2015; and Considine, G & Zappalą, G 2001, Factors Influencing the Educational Performance of Students from Disadvantaged Backgrounds, Accessed 20 August 2015 4. Australian Bureau of Statistics (ABS) 2015, Educational outcomes, experimental estimates, Queensland, 2011 - Socioeconomic factors and early childhood development in Queensland, cat. no. 4361.3, Accessed 17 August 2015

5. AEDC, The AEDC domains, Accessed 17 August 2015
6. AEDC, FAQ for researchers, Accessed 17 August 2015
7. ABS, Preschool Education, Australia, 2014 (see Background information), cat. no. 4240.0, Accessed 7 September 2015
8. ABS, Preschool Education, Australia, 2012 (see Table 23 in Download tab), cat. no. 4240.0, Accessed 7 September 2015
9. Australian Government, A Snapshot of Early Childhood Development in Australia 2012 - AEDI National Report Re-issue November 2013, p. 3, Accessed 18 August 2015
10. See end note 9.
11. Tasmania Department of Education, Admission Guidelines for Kindergarten and Preparatory (Prep), p.2, Accessed 18 August 2015
12. 'BetterStart' Child Health and Development Research Group 2014, Five by Five: A Supporting Systems Framework for Child Health and Development, p. 6, Accessed 18 August 2015
13. Miranti, R, Daly, A & Tanton, R 2015, 'An area-based measure of risk of social exclusion for Australian school-age children', Australasian Journal of Regional Studies, pp. 26-49
14. Salt, B, Tree-changers boosting inland Tasmania, Accessed 18 August 2015
15. Productivity Commission, Report on Government Services, 2011, p. 3.28, Accessed 18 August 2015
16. ABS 2014, Educational outcomes, experimental estimates, Queensland, 2011 and Educational outcomes, experimental estimates, Tasmania, 2006-2013, Accessed 18 August 2015