Growing Up in Australia and Footprints in Time: The LSAC and LSIC Research Conference, 2011

Abstracts

Professor Steve Zubrick, Head, Division of Population Science, Telethon Institute of Child Health Research

‘Digital Natives'? The effects of new and old media on children's language acquisition and literacy

Michael Bittman, Leonie Rutherford and Jude Brown, School of Behavioural, Cognitive and Social Sciences, University of New England

The current generation of young children has been described as ‘digital natives', having been born into a ubiquitous digital media environment. They are envisaged as educationally independent of the guided interaction provided by “digital immigrants”: parents and teachers. This paper uses data from the multiple waves the Longitudinal Study of Australian Children (LSAC) to study the effect of various media on children's development of vocabulary and traditional literacy. Previous research has suggested that time spent watching television is associated with less time spent reading and, ultimately, with inferior educational outcomes. The early studies of the ‘new', digital media (computers, games consoles, mobile phones, the internet, etc.) assumed these devices would have similar effects on literacy outcomes to those associated with television. Moreover, these earlier studies relied on poorer measures of time spent in media use and usually did not control for the context of the child's media use. Fortunately LSAC contains measures of access to digital devices; parental mediation practices; the child's use of digital devices as recorded in time-diaries; direct measures of the child's passive vocabulary; and teachers' rating of the child's literacy. The analysis presented shows the importance of the parental context framing the child's media use in promoting acquisition of vocabulary, and suggests that computer (but not games) use is associated with more developed language skills. Independently of these factors raw exposure to television is not harmful to learning as previously thought.

Footprints in Time: Learning language, learning culture

Laura Bennetts Kneebone, Department of Families, Housing, Community Services and Indigenous Affairs (FaHCSIA)

Opinions and stereotypes surrounding Indigenous languages are widespread, but what do the linguistic environments of Australia's Aboriginal and Torres Strait Islander children really look like? Are Indigenous languages still actively spoken in the home? What about in homes where one parent is non-Indigenous? How are families who have maintained their languages similar or different to families who haven't? Bilingual families have a different set of strengths and challenges to monolingual families in deciding what they want for their children and this presentation will explore what these attitudes and choices are and look at their relationship to babies' vocabulary acquisition. The MacArthur-Bates Communicative Development Inventory is an instrument used by the Longitudinal Study of Indigenous Children to measure babies' English vocabulary development at Waves 1, 2 and 3.

In Wave 3, 99.9 per cent of LSIC families spoke at least some English in the home. Over a quarter spoke more than one language, of these 71 per cent said that they speak English often or a lot in the home. This leaves just under a third of bilingual families who speak little or no English in the home. This paints a picture of Indigenous families as predominantly English speaking monolinguals, but scratching a little more deeply beneath the surface, nearly 50 per cent of families reported speaking some form of Aboriginal English.

The data shows a number of cultural differences between Aboriginal English (AE) speaking families and Standard Australian English (SAE) speaking families. For example, AE speaking families placed a much higher priority on passing down to their children Indigenous cultural practices relating to bush tucker, hunting and fishing; tradition and ceremony; and speaking language. SAE speaking families were more likely to emphasise family history; pride in identity and family networks. The same trends can also be seen when contrasting remote and urban families, but there is not a one-to-one correspondence between speaking Aboriginal English and location. Although families living in remote areas were highly likely to speak Aboriginal English, there were lots of families in urban and regional areas who identify as speaking it too.

Early vocabulary development: The importance of joint attention and parent-child book reading

Brad Farrant and Stephen Zubrick, Telethon Institute for Child Health Research, The University of Western Australia

A large body of work has investigated the individual and environmental factors associated with children's language development. However, there is a dearth of research assessing the relations among these factors and the ongoing reciprocal social interactions (proximal processes) that facilitate development. The current study brought a bioecological approach to children's early vocabulary development using data from the B cohort of the Longitudinal Study of Australian Children. Relevant longitudinal data were available for 2188 children (1119 male) who had a median age of 9 months (M = 9.3 months, SD = 2.1 months) at Wave 1 and a median age of 34 months (M = 34.2 months, SD = 2.5 months) at Wave 2. Results indicate that the proximal processes of joint attention and parent-child book reading facilitate vocabulary development and mediate the effects of maternal education, warm parenting, and having siblings in the home. Furthermore, combining the current results with the findings of training studies provides convergent support for the causal roles of joint attention and parent-child book reading in children's vocabulary development. Thus, parents and professionals interested in promoting children's early language development would be well served by targeting proximal processes like joint attention and parent-child book reading especially when working with disadvantaged children.

The development of young children of immigrants in Australia, Canada, the United Kingdom and the United States

Elizabeth Washbrook, University of Bristol, Jane Waldfogel, Columbia University & London School of Economics, Bruce Bradbury, University of New South Wales, Miles Corak, University of Ottawa and IZA and Ali Akbar Ghanghro, University of Ottawa

This study examines the developmental outcomes of 4-5 year old children with immigrant parents. This comparison uses data from the first three waves of the Australian LSAC child cohort, together with similar datasets in Canada, the United Kingdom and the United States.

While all four of these countries have historically been immigrant receiving nations with common cultural, demographic and labour market characterisitcs, the selection rules they use to screen immigrants and the support offered to them differ markedly. At one extreme Australia places a strong emphasis on choosing immigrants for their education and language skills, characteristics that are likely to encourage labour market success. At the other extreme immigrants to the United States are in large measure chosen for reasons of family re-unification, with a substantial fraction also being undocumented migrants with miminal education and English language skills.

Our research suggests that these selection mechansims geared to labour market outcomes do not lead to differences in the capacities of families and parents to give their children a reasonable start in life. We find that the four and five year old children of immigrant families have the cognitive and behavioral skills required to successfully start school in all of these countries. This contrasts with findings from other research focusing on outcomes during the teen years, which document considerable cross-country differences. Teenagers of immigrants in Canada and Australia perform as well as, or even better, than children of the native born in reading, math and science tests; in the United Kingdom and the United States they on average perform worse.

Has the parental education – child outcome gradient narrowed in Australia since the 1980s?

Gerry Redmond Social Policy Research Centre, Bina Gubhaju, Australian National University, Diana Smart, Australian Institute of Family Studies and Ilan Katz, Social Policy Research Centre

The purpose of this paper is to examine how the relationship between parents' educational achievement, a marker of their socio-economic status, and children's early developmental outcomes has evolved in Australia since the early 1980s.

We specifically focus on the question of whether there is any evidence that the relationship between parents' education and children's developmental outcomes has changed during the past three decades. Our study builds on work by Smart and Sanson (2005) who found that children's temperament and behaviour in the ATP and LSAC surveys are similar. The small but significant shifts that have occurred point to the LSAC children doing slightly better than the ATP children. Smart and Sanson measured outcomes for the whole cohorts and did not study the differences between different sub-populations.

We examine this question through a comparative analysis of the Australian Temperament Project (ATP) and the first three waves of the LSAC K cohort. We analyse the relationships between parents' education and children's developmental outcomes in two functional domains of child development – social/emotional, and cognitive. We also examine whether there have been changes in the extent to which children's persistently high or low levels of functioning between the ages of about 4 and 11 (ATP) or 9 (LSAC), is related to their parents' educational achievements.

The overall ‘big picture' finding of this study is that parental education is strongly associated with children's cognitive and socio-emotional wellbeing in both cohorts. With regard to changes over time our analysis indicates that although there has been a significant improvement in the levels of parental education since the 1980s, this has not resulted in a narrowing of differences between the wellbeing of children from different backgrounds; if anything the gradient between parental education and children's outcomes has steepened over the past three decades. However there are some differences in the mediating variables between parental education and children's early outcomes, which indicate that changes in educational practices may result in better outcomes for more troubled children. The findings also raise questions about the nature of the relationship between parental education and child wellbeing, which have significant policy implications.

A study of Indigenous children's developmental outcomes: the impact of child, family and socio-economic characteristics

Killian Mullan and Gerry Redmond, Social Policy Research Centre, University of New South Wales

Closing the gap in early child development between Indigenous and non-Indigenous children is a core aim of Federal, State and Territory Governments. Addressing this aim raises a complex set of issues and yet, all too often, Indigenous people and families are simply regarded as a single unitary group. This is unfortunate, as efforts to close the gap stand a much better chance of success if they are based on a more nuanced understanding of the relationship between disadvantage and child development within Indigenous communities. This paper offers a small step towards a better understanding of the relationship between child development outcomes and characteristics associated with the child, family, the broader environment and socio-economic conditions. Using data from the first wave of the Longitudinal Study of Indigenous Children we present a detailed picture of the characteristics of Indigenous children, their primary carers, their environment, and aspects of the socio-economic conditions within which they develop. Focusing on Indigenous children 3-6 years old, we look at associations between these characteristics and outcomes relating to children's readiness to learn, their early language development, and their socio-emotional development. Our results show a complex set of relationships and we discuss implications for policy development and future research.

The assessment of temperament in 4 to 5 year old Indigenous children

Keriann Little, Department of Paediatrics, The University of Melbourne, Stephen Zubrick, Telethon Institute for Child Health Research and Ann Sanson, Department of Paediatrics, The University of Melbourne

Currently little is known about the structure of early temperament or its influence on emotional and behavioural adjustment in Aboriginal and Torres Strait Islander children in Australia. This paper will provide psychometric information on measures of temperament from the Longitudinal Study of Indigenous Children (LSIC), when participants were aged 4 1/2 - 5 1/2 years old. The paper will also provide insight on these children's temperament profiles, and the dynamics that exist between temperament and other factors. We seek to address three key questions: 1) What is the structure and nature of temperament in young Indigenous children and how does this compare to non-Indigenous children? 2) Does the temperament of Indigenous children vary according to factors such as gender and remoteness? 3) Do associations exist between early temperament, parenting practices and later emotional and behavioural difficulties in Indigenous children? Analyses will draw on data from waves 1-3 of the Longitudinal Study of Indigenous Children (LSIC) as well as Indigenous children from the Longitudinal Study of Australian Children (LSAC), with a group of non-Indigenous children in LSAC used as a comparison group. Given that temperament is regarded as largely constitutionally based, it is predicted that no differences will be found in temperament structure or profiles between these groups. It is expected that more “difficult” temperament characteristics, such as shyness, emotional reactivity and poor attention, and less positive parenting practices will be associated with more emotional and behavioural problems one year later. In investigating this topic, this paper will expand the current evidence on what factors promote positive outcomes for Indigenous Australian children.

Child temperament, parenting style and emotional and behavioural problems in early childhood - findings from the Longitudinal Study of Australian Children

Daniel Christensen, Telethon Institute for Child Health Research, Francis Mitrou, Telethon Institute for Child Health Research, David Lawrence, Telethon Institute for Child Health Research, Ann Sanson, Department of Paediatrics, University of Melbourne and Stephen R Zubrick, Telethon Institute for Child Health Research

Identifying and describing the antecedents of emotional and behavioural problems in early childhood has been of much interest to researchers in the field of child behaviour. Recent studies have highlighted the diagnostic validity, persistence and sequelae of these problems. Within this broad research area there is particular interest in possible interactions between parenting styles and child temperament, such as children with particular temperament characteristics exposed to less favourable parenting practices are disproportionately at risk of emotional and behavioural problems. This has been described as the ‘goodness of fit' model. Research with children and adolescents provides support for the impact of parenting and temperament on emotional and behavioural problems, as well as inconsistent support for the interactions predicted by the goodness of fit model.

This study sought to add to the current knowledge in the field of early childhood behaviour by examining the goodness of fit model in a group of 2,644–3,214 children from the B cohort of the Longitudinal Study of Australian Children. The temperamental characteristics of approach and irritability were measured at 3–19 months, parental warmth and hostility were measured at 2–3 years, and emotional problems, peer problems, conduct problems and hyperactivity were measured at 4–5 years via parent-report. Using logistic regression models, we found that child temperament and parenting style predicted the subsequent development of emotional and behavioural problems, independent of a range of other biopsychosocial ecological factors. However, we did not find the hypothesized interactions between parenting and temperament. On the basis of the particular temperament and parenting characteristics examined here, our results suggest that goodness of fit models may not accurately describe the contribution of parenting and temperament to the development of emotional and behavioural problems in early childhood.

Differential susceptibility to the environment: An examination of the moderating effect of temperament on the relation between the home environment and child development

Michelle Ernst, Linda Collins and Fred Wulczyn, Chapin Hall at the University of Chicago

The theory of differential susceptibility to the environment suggests that children differ in how greatly they are impacted by the environment as a function of their temperament. According to the theory, the relation between environment and outcome is moderated by children's temperaments such that children with difficult temperaments are greatly benefited by a positive environment but greatly disadvantaged by a poor environment. By contrast, children with easy temperaments tend to be affected much less by their environment – for good or bad.

This area of research is thought to have the potential to help us better understand the interaction between “nature” and “nurture” as well as to inform interventions. Studies have already found evidence that child temperament plays a moderating role between child outcomes and a variety of environmental influences such as discipline, childcare, and attachment.

In this paper we examine this theory using data from the Longitudinal Study of Australian Children (LSAC). The LSAC data is uniquely positioned to address these questions as it provides longitudinal data that captures the moderating variable (temperament) temporally prior to the developmental outcomes of interest. Temperament is measured in the first year of life for many respondents while data capturing developmental outcomes are collected when children are aged 4-5 years. Information about the home environment is collected at all three waves of data collection.

Using structural equation modelling, we examine the relation between the home environment and developmental outcomes as a function of a child's temperament. Outcomes will be defined as social and cognitive development at wave 3 when children are 4-5 years of age. We test the hypothesis that children with difficult temperaments exhibit a strong relation between the quality of their home environment and their social and cognitive development at 4-5 years of age.

Footprints in the Field

Ros Thorne, Sharon Barnes and Lu Hidderley, Department of Families, Housing, Community Services and Indigenous Affairs (FaHCSIA)

Footprints in Time: The Longitudinal Study of Indigenous Children (LSIC) was funded in 2003-04 and incorporated two years of community engagement across Australia and two years of testing and piloting. The study reflects the input from Aboriginal and Torres Strait Islander communities in its survey design and content including: a community based design; a focus on positive outcomes and resilience; and a commitment to ongoing feedback.

How does the design and operation of Footprints in Time support these commitments and build relationships and trust with participants? Aboriginal and Torres Strait Islander Research Administration Officers (RAOs) conduct the parent/carer and study child interviews for the study, primarily in their own communities. RAOs are employed by the Department of Families, Housing, Community Services and Indigenous Affairs (FaHCSIA) in full time, permanent positions involving interviewing and community liaison.

The RAOs have been responsible for participants remaining engaged, informed and willing to participate. Their presence in communities, collecting data on a yearly basis, has meant that strong and trusting relationships have been built with the families involved in the study. This is shown in one of the key successes of LSIC, which is its retention rate of 86% over the first three waves of the study.

A demonstration of the Study Child questionnaire will also be provided, to highlight the innovative solutions implemented by the design team.

Data collection instruments – Wave 5 of the Longitudinal Study of Australian Children

Australian Bureau of Statistics LSAC team

As the children in our study have grown and changed, so have the methods we use to collect the data. We change our methods of collection for a number of reasons:

To promote engagement in the study by our respondents
To keep up-to-date with technology
To improve data quality and response rates
To decrease processing time

This led us to making significant changes to the Wave 4 data collection procedures. These changes were successful and Wave 5 will continue with the same instruments, updated with age appropriate content, and design improvements.

Wave 5 will utilise the following parent data collection instruments:

A computer assisted telephone interview (CATI)
A computer assisted self interview (CASI)
An Interviewer administered computer assisted interview (CAI)
A self-complete paper questionnaire

Wave 5 will also employ the following child data collection instruments:

An Interviewers administered child self report
A number of different cognitive assessments
A combined paper/electronic time use diary
An audio computer assisted self interview
An accelerometer

This presentation will describe the different data collection methodologies developed for the recent Wave 5 Dress Rehearsal. Demonstrations of all instruments will be available.

Stability and change in teacher-child relationships trajectories from age 4-5 to 8-9 years: Associations with academic achievement in Year 5

Linda Harrison, School of Teacher Education, Charles Sturt University, Janine Spilt, Department of Developmental Psychology, VU University and Susan Walker, Faculty of Education, Queensland University of Technology

The importance of supportive, non-conflicted relationships between students and teachers for academic achievement and school adjustment has been well established in Australia and internationally, but few studies have examined these effects over time. In this paper we draw on four waves of the LSAC-K cohort data to test key questions about teacher-child relationships and educational outcomes: (1) are teacher-student relationships stable or changeable over time? (2) do relationship trajectories differ by child gender?; and (3) do different trajectories differentially predict achievement outcomes?

We draw on the work of Spilt et al. (in press) who identified four latent class growth trajectories for teacher-student conflict from Grade 1 to 5: low-stable (normative), low-increasing, high-declining, and high-stable (only for boys), and two for teacher-student closeness: high-declining (normative) and low-increasing in a U.S. sample of low-achieving children. We test these findings in the more representative sample provided by LSAC, using growth mixture models for teacher-reported closeness and conflict in the K-Cohort at ages 4-5, 6-7 and 8-9 years, conducted separately for boys and girls.

Results confirmed the normative patterns reported by Spilt et al. For teacher-child conflict, 73% of boys and 85% of girls were identified as having a low-stable growth trajectory. For closeness, 56% of boys and 63% girls were identified as having a high-declining trajectory. Five additional latent class trajectories were identified for conflict, with some sex-related differences: for girls, low increasing (9%); high declining (4%); and moderate/high-increasing (2%); and for boys, low-increasing (12%); high declining (8%); low/moderate-increasing (4%); and high stable (3%). In contrast, boys' and girls' non-normative trajectories for closeness were similar in proportion, intercept, and slope: high stable (30% boys; 28% girls); low increasing (11% boys; 7% girls); and low increasing (3% boys; 2% girls).

With the availability of Wave 4 K Cohort NAPLAN scores, further tests will be conducted to examine trajectories in teacher-child relationship conflict and closeness on academic achievement, controlling for family socio-economic position, child gender, ethnicity and early literacy/numeracy ability assessed at age 4-5 years.

Resilience and educational outcomes in Footprints in Time

Annette Neuendorf, Department of Families, Housing, Communities and Indigenous Affairs (FaHCSIA)

The first three waves of parent and study child data will be explored in this presentation. It will look at Wave 3 outcome data for the child cohort: the Strengths and Difficulties Questionnaire (SDQ), Who Am I?, Renfrew and teacher rated literacy scores and examines what affects these scores.

What can LSIC tell us about the experiences Indigenous children and parents with preschool and school? Initial analyses suggest that children in Wave 2 who had been to preschool in Wave 1 tended to score higher on measures of vocabulary and school readiness than children who had not been to preschool or childcare. Initial analyses also suggest most children in the study do like school, are not bullied and are attending every day. The first three waves of parent and study child data will be explored in this presentation which provides unique insights into the educational experiences of Indigenous children and parents.

With parent or guardian permission Footprints in Time sends questionnaires to the teachers and carers of the children in the study. This presentation will also discuss some of the preliminary findings from the teachers' surveys, such as: the strengths of the study children that teachers identify, how far teachers predict a child will go in their education and the kinds of Indigenous specific activities and training available within their schools.

Early Pathways and Capabilities

Nicola Taylor, Department of Education, Employment and Workplace Relations (DEEWR)

While there is considerable evidence of individual, family and socio-economic disadvantage for children's school readiness and subsequent developmental outcomes, there is little research on mediating mechanisms in the early years of schooling. These mechanisms are particularly important as potential policy levers to facilitate better outcomes for children who experience early disadvantage.

The research has three objectives. First we examine the extent to which initial conditions set trajectories in the transition to schooling. Using wave 1 data of the LSAC kindergarten cohort, we identify early disadvantage associated with low school readiness in five contexts: individual factors, socio-economic factors, parenting characteristics, home learning environment and location. Second, based on the elements of disadvantage that appear to be most detrimental in school readiness, we construct cumulative risk indexes to demonstrate the relationship between higher disadvantage in one context and higher disadvantage in other contexts. Third, we use logistic regressions to trace the link between these initial conditions (risk indexes) and children's cognitive, learning and wellbeing outcomes at age 8-9 years, to answer the key research question of whether there are any mediating mechanisms at waves 2 and 3 that reduce the impacts of initial conditions.

Outcome measures at wave 3 are PPVT scores (cognitive development), Marsh self-perception scores (wellbeing) and literacy scores (achievement).

Controlling for cognitive ability, school readiness and other factors we find that a number of behaviours and traits at waves 2 and 3 are related to wave 3 outcomes. Persistence, socio-emotional problems and parental expectations about education are important predictors of literacy achievement, reducing the impacts of early disadvantage. For wellbeing there are gender differences, with positive mediators such as pro-social behaviour and persistence even eliminating the negative effects of early parenting disadvantage for boys' wellbeing, but not for girls'. Negative factors like stealing, aggression and socio-emotional problems reduce wellbeing in both sexes. Cognitive development is highly related to early disadvantage but largely resistant to later mediation, confirming the importance of early intervention for this outcome.

The effects of early paternal depression on children's development

Richard Fletcher and Emily Freeman, Family Action Centre, University of Newcastle , Craig Garfield, Departments of Pediatrics and Medical Social Science, Northwestern University, Chicago and Graham Vimpani, School of Medicine and Public Health, University of Newcastle

Infants depend on parents to develop their cognitive and emotional capacities. Internationally, maternal postnatal depression has been shown to adversely affect child wellbeing and development, so much so that maternal screening is standard in many countries. However, more recently, evidence suggesting a causal relationship between fathers' postnatal depression and children's disrupted development has emerged using a large UK sample. The aim of the current study was to examine the effects of paternal depression during the child's first year on their wellbeing at 4-to-5 years using a large, representative sample of Australian families.

Two-biological-parent families (N=2620) from the LSAC B-cohort were included if depression measures were available for both parents in 2004, behavioural and developmental measures were available for children in 2008, and the families had not separated by 2008. Logistic regression modelling was used to investigate how early paternal depression in 2004 influenced child outcomes 4-years later. Early paternal depression was found to be a significant predictor of a range of poorer child outcomes and these effects remained significant after controlling for both early maternal depression and later paternal depression.

Our study shows that depression in fathers during the first year of a child's life can have a detrimental impact on their child's behavioural, social and emotional development at the point of school entry, in addition to and uniquely compared to mothers. These findings suggest that early intervention to identify and address the mental health needs of fathers is required for the benefit of fathers, children, and families.

True blue baby blues: Prevalence and persistence of mental health difficulties and poor socio-emotional wellbeing in Indigenous Australian fathers with infants and young children

Fabrizio D'Esposito, Rebecca Giallo, Paul Stewart, Fiona Mensah and Jan Nicholson, Parenting Research Centre

Background: Paternal mental health and socio-emotional wellbeing (MHSEWB) are important determinants of children's short- and long-term development and wellbeing. Research in non-Indigenous populations shows that fathers of infants are at greater risk of experiencing mental health difficulties than other men. These problems may persist into the early childhood period. Data on the MHSEWB of Indigenous fathers is lacking, however mental health problems have been recognised as a major difficulty for most Indigenous communities.

Aim: The main aim of this study is to describe the prevalence and persistence of poor MHSEWB in Indigenous fathers with infants and young children. A secondary aim of the study is to identify factors associated with poor MHSEWB in Indigenous fathers during the first postnatal year. Methods: This study will use data from waves one and two of the “Baby cohort” of Footprints in Time: the Longitudinal Study of Indigenous Children. The main outcome will be MHSEWB as assessed by the “Social and Emotional wellbeing” section of the study parent questionnaire. Descriptive analyses will be used to provide socio-demographic information on the sample population and to investigate the prevalence and persistence of poor MHSEWB among Indigenous fathers. Bivariate and multi-variable logistic regression analyses will be used to investigate associations between socio-demographic, individual, contextual, and cultural factors and poor MHSEWB in Indigenous fathers during the first postnatal year.

Results and conclusions: For the first time we will report on the prevalence of poor MHSEWB in Indigenous fathers with infants and young children. The persistence of paternal MHSEWB problems beyond the post-partum period and into the early childhood period will also be reported. Furthermore we will report on a range of risk factors that may predict poor MHSEWB in Indigenous fathers. It is likely that findings will provide guidance to better direct policy efforts to address needs that are specific to Indigenous fathers and which differ from those of their non-Indigenous counterparts.

Fathers' postnatal mental health and child wellbeing at aged five: The mediating role of parental self-efficacy and parenting behaviour

Rebecca Giallo, Amanda Cooklin, Catherine Wade,Fabrizio D'Esposito and Jan Nicholson,Parenting Research Centre, Murdoch Children's Research Institute & RMIT University

Background: Fathers' postnatal mental health difficulties have been associated with both short- and long-term emotional and behavioural outcomes for children. The mechanisms or pathways by which fathers' mental health may contribute to outcomes for children have received little attention. One pathway is likely to be via the negative effects of mental health difficulties on parenting behaviours important for promoting child wellbeing and development, however, these associations have yet to be substantiated longitudinally or with fathers.

Aim: To explore the longitudinal relationships between fathers' postnatal mental health, parenting and emotional-behavioural outcomes for children at aged five.

Study Design: Secondary analysis of data from fathers participating in the Growing Up in Australia: the Longitudinal Study of Australian Children was conducted. Participants were 2025 fathers in a couple relationship who had complete data on measures of psychological distress, parental self-efficacy, parenting warmth and irritability. Mothers' reports of children's emotional and behavioural functioning were used. Data were from three waves of data collection when the children were aged 0-1, 2-3 and 4-5 years.

Results: Path analysis and latent growth modelling revealed that the relationship between fathers' postnatal psychological distress and emotional-behavioural outcomes for children at age five was mediated by parental self-efficacy (PSE) and parenting behaviour. Specifically, fathers' psychological distress was associated with low PSE in the postnatal period and a decline across the early childhood period. Low postnatal PSE and over time was in turn predictive of low parenting warmth and high parenting irritability when the children were aged five. Low postnatal PSE and decline over time, low warmth and high irritability were associated with increased child emotional and behavioural difficulties after controlling for concurrent fathers' mental health. These relationships were also maintained after accounting for the relationship between mothers' postnatal mental health and child outcomes.

Employment and family conflict: parenting quality and couple relationship for mothers of 4-5 year olds

Amanda Cooklin, Parenting Research Centre, Elizabeth Westrupp, Murdoch Children's Research Institute, Lyndall Strazdins^,National Centre for Epidemiology and Population Health and Jan Nicholson^,Parenting Research Centre

Background: The aim of this study was to investigate the relationship between employment strains and gains, parenting and the couple relationship for mothers of 4-5 year old children.

Method: Data were from mothers of 4-5 year old children at Wave 1 (K Cohort, N=2115). Employment strains and gains were assessed using 10-items adapted from Marshall and Barnett (1993). Primary outcomes were four parenting variables (Irritability; Warmth; Consistency; Parental Self-Efficacy), and two assessments of the couple relationship, Couple argumentativeness; Relationship quality. We used unadjusted and adjusted multivariable linear regression to assess the relationship of employment strains and employment gains to parenting and couple functioning. Adjusted analyses controlled for maternal and infant characteristics.

Results: Higher employment strains were significantly associated with lower warmth ? (95% CI) = -0.03 (-0.06, -0.01); increased argumentativeness ? (95% CI) = -0.05 (0.02, 0.07); and poorer quality couple relationship ? (95% CI) = -0.05 (-0.09, -0.01) when all maternal and infant characteristics were adjusted for. Employment gains were associated with higher parenting warmth ? (95% CI) = 0.03(0.01,0.06), increased parenting consistency ? (95% CI) = 0.06(0.01, 0.09) and better quality couple relationship ? (95% CI) = 0.10 (0.04, 0.06) in adjusted analyses.

Conclusion: Employment strains and gains were shown to influence parenting behaviours and the couple relationship. Mothers' employment strains are associated with poorer quality couple and child relationship, and both are important for child outcomes. Identifying pathways via which workplaces, jobs and employments effect families lends further weight to policy provision of high quality, family-friendly optimal workplace environments.

Investigating stability and change in multiple childhood adversity over time

Bryan Rodgers, Bina Gubhaju, Peter Butterworth, Lyndall Strazdins, Tanya Caldwell and Tim Crosier. Australian Demographic and Social Research Institute (ADSRI), Australian National University

The main aim of this presentation is to examine the persistence of multiple adversity over time. Past research has focused on specific aspects of childhood adversity, typically by controlling or adjusting for other factors when examining the association between a particular risk factor and a particular outcome. The present project is, instead, concerned with the experience of multiple disadvantage. As part of this research, we have developed summary measures of childhood adversity from the Longitudinal Study of Australian Children (LSAC) data over the first three waves. This presentation reports on the persistence of adversity over time, using measures that cover:

(1) individual indicators (i.e. items); (2) construct measures covering 12 areas of adversity (family structure, economic position, financial hardship, employment conditions, stressful life events, parental substance use, parent wellbeing, time pressure (general), time pressure (work), parenting, parent-parent relationships, social support); (3) higher-order measures representing two components labelled socio-economic and psychosocial adversity;

(4) total adversity scored across all the constructs covered by the study.

This range of measures allows a number of analytical approaches that inform our scientific understanding of the nature of childhood adversity. For example, comparison across the various types of derived measures outlined above will determine whether higher-order measures show greater continuity over time than is found for specific indicators. Other multivariate analyses will indicate what combinations of measures at an earlier stage of the life-course are the best predictors of adversity at a later point. Further analyses address more particular policy relevant questions. As an illustration, we can focus on families who experience particular transitions, such as the shift from joblessness to being a “working family” and determine whether such transitions are accompanied by improvements across other measures of adversity or whether these families remain disadvantaged in other respects in spite of the transition to having paid employment. Similar analyses can be conducted for other policy relevant transitions, such as from two-parent to sole-parent families and vice versa.

This project contributes to advancing methodologies that attempt to quantify the nature and persistence of simultaneous occurrences of various forms of disadvantage. It will also aid in policy development by identifying families that consistently experience multiple forms of adversity.

Why some children thrive despite the odds being stacked against them: Investigating the characteristics of young mothers and their children who have good outcomes

Helen Rogers, Helene Shin and Rajinder Ghooti, Department of Families, Housing, Community Services and Indigenous Affairs (FaHCSIA)

A number of Australian government policies and programs are designed to help improve the wellbeing of young mothers and their children. However, there is a growing body of child development literature suggesting that children born to young mothers have poorer outcomes compared to children born to older mothers. Although findings from national and international studies vary it has been reported that children of young mothers are more likely to have socio-emotional problems, poorer school performance and leave school at an earlier age than those born to older mothers. Furthermore, they face greater risk of alcohol and drug abuse, living in poverty, being involved in crime, becoming teenage parents themselves and being married and divorced earlier.

Although the negative outcomes of being born to young mothers are well documented, few Australian studies have explored the protective factors that enable children of young mothers to thrive.

The present study examines the characteristics of young mothers (e.g., socio-demographic attributes, employment status, mental-emotional wellbeing) and the environment in which the child is growing up (e.g. family structure, housing, schooling, activities, and resources) to investigate the consequent outcomes for children. This will provide an understanding of why some children with young mothers do thrive, despite the potential problems they face. To investigate this relationship, data from the Longitudinal Study of Australian Children (LSAC) was analysed using a cross-sequential design.

Findings of the study highlight factors that shield children born to young mothers from negative outcomes and identify key issues to inform policy decisions concerned with improving outcomes for young parents and their children. Further analyses and policy implications will be discussed in detail during the presentation.

Keynote Presentation

The Health of Australia's Children

Professor Melissa Wake, Director of Research, Centre for Community Child Health, Royal Women's Hospital

Does childcare in the first year of life pose a risk for concurrent and future poorer childhood health?

Linda Harrison, School of Teacher Education, Charles Sturt University, Brad Farrant, Telethon Institute for Child Health Research, Sarah Wise, Policy Research and Innovation Unit, Anglicare Victoria and Grant Smith and Stephen Zubrick, Telethon Institute for Child Health Research

This paper seeks to understand the role of childcare attendance during infancy on concurrent and future risk of infectious illness. Previous research comparing children receiving child care with those in exclusive parental care generally conclude that child care predisposes children to more episodes of upper and lower respiratory tract infections, a longer duration of each illness, and increased risk in the first two years of life. Longitudinal studies suggest a trend toward stabilized or decreased rates of infection over time; however, this process is less well understood. Some authors suggest that early exposure in group care ‘inoculates' children and reduces later susceptibility to illness. Others argue there is no evidence that children with many recurrent infections become more resistant to infection with increasing age, citing evidence that recurring middle ear and respiratory infections in childhood may have chronic sequelae.

The LSAC Birth cohort data provide an opportunity to investigate this issue in Australia. To address the question of whether early child care is related to concurrent and future risk for illness, wave 1 data were examined to identify families who had provided information about their use of child care in the first year of life. A sample of 3264 families was identified, of whom 2081 did not use childcare prior to 12 months and 1543 did. Type of care attended in the first year was summarized as: formal (long day care centre or family day care home), informal home-based care (e.g., with grandparent), and mixed formal and informal care. Amount of care was coded as hours of centre care, hours of family day care, and hours of informal care. Illness was coded as a binary variable (has ongoing problems with ear infections; no problems).

The research questions were tested in a series of logistic regression analyses, controlling for sex of child and eleven demographic/protective/risk factors. Risk for concurrent illness was tested by regressing type of care and amount of care prior to 12 months on infections reported at wave 1. Risk for future illness was tested by regressing type and amount of care prior to 12 months on infections reported at waves 2 and 3, controlling for current type and amount of care.

Results showed a consistent pattern that aligned with the findings reported in the literature. At each wave of data, concurrent infections were significantly higher in children attending formal group care settings than in children in exclusive parental care. The highest rates were for children attending more than 30 hours per week of centre care. Informal care was not associated with a rise in infections. Longitudinal analyses, however, showed no relationship between type and amount of early childcare and rates of infection at wave 2 or 3. Significant predictors were child prematurity, mothers having less than a Year 12 education, and the presence of older siblings in the household. Possible processes explaining childhood health and early childhood settings will be discussed.

Playgroups as sources of social support for mothers

Kirsten Hancock, David Lawrence, David Zarb, Francis Mitrou, Donna Berthelsen, Jan Nicholson and Stephen Zubrick, Telethon Institute for Child Health Research, The University of Western Australia

Playgroups are regular, informal gatherings of parents and young children that provide a range of developmental opportunities for children, as well as opportunities for parents to develop social and support networks. While recent research has demonstrated that children from disadvantaged families tend to perform better on measures of learning and social outcomes when they persistently attend playgroup (Hancock et al., 2011), there is still very little research demonstrating any association between playgroup attendance and parental outcomes. The purpose of this study therefore was to examine the friendship networks and social support outcomes of mothers of young children according to patterns of playgroup attendance over time. This study used data from the B-cohort of the Longitudinal Study of Australian Children across three waves to examine the level of support that mothers received from friends over time. Multinomial logistic regression modelling indicated that mothers of 4-5 year olds who participated in playgroups when their child was 0-1 and 2-3 years were more likely to have consistently good support from friends, or to have improved support from friends compared with mothers who did not participate in playgroups with the study child. These results provide some evidence that playgroups are associated with improved social networks over time, and that socially isolated parents may find playgroups a useful resource to build their social networks.

Preschool Participation among Indigenous Children

Belinda Hewitt, Institute for Social Science Research, University of Queensland and Maggie Walter, Department of Sociology and Social Work, University of Tasmania.

In this paper we examine the structural, demographic and cultural factors that are associated with preschool attendance amongst Indigenous children. It is well documented that a smaller proportion of Indigenous four and five year olds attended pre-school compared with same age non-Indigenous children. Furthermore, the participation gap increases with level of remoteness. Yet, the (mostly overseas) research indicates that early childhood years are extremely important and formative in setting the framework for later educational achievement and that the benefits are magnified for those from disadvantaged backgrounds. This suggests that poorer preschool attendance further disadvantages Australian Indigenous children, who already face substantial barriers to their life chances. The preschool attendance gap is of major policy concern, evidenced by the Closing the Gap measure of ensuring all Indigenous four year olds in remote communities have access to early childhood education within five years (by 2013).

Despite the social and policy importance, very little empirical research to date has investigated the factors associated with preschool attendance amongst Australian Indigenous children. To investigate this issue we use the first two waves (2008 & 2009) of the Longitudinal Study of Indigenous Children (LSIC). Our analytic sample includes 670 children who were eligible to attend preschool from the child cohort. We limited the sample to children who were aged 4 years or older in Wave 1 and/or Wave 2 and not attending their first year of formal schooling; there 124 children who were eligible for pre-school in both years. We use a 2-level multi-level model that controls for the area-clustered design of the study and the longitudinal nature of the data. We find that a range of factors influence preschool attendance amongst this group of children. As study children get older the likelihood of them attending preschool increases. We also find that if the primary source of household income is derived from government benefits that the study child is less likely to be attending preschool. Finally, study children who attend cultural events with their families are less likely to attend preschool.

Time trends in the incidence and prevalence of asthma in Australian children: a cohort and age-period analysis

Rosario Ampon, Leanne Poulos, Stephanie Cooper, Anne–Marie Waters, Helen Reddel and Guy Marks, Australian Centre for Asthma Monitoring (ACAM), Woolcock Institute of Medical Research

Background: Knowledge about the incidence and prevalence of asthma in children can help researchers and policy-makers predict demand for health care.

Objective: To examine time-trends in the cumulative incidence of doctor-diagnosed asthma and the prevalence of current asthma using data from two cohorts of the Longitudinal Study of Australian Children (LSAC).

Methods: We estimated the cumulative incidence of parent-reported ever-diagnosed asthma and the prevalence of current asthma in the infant (aged 0-1 year at baseline) and kindergarten (aged 4-5 years at baseline) cohorts at each of two follow up assessments. Cumulative incidence was defined as having reported ever-diagnosed asthma at the current or any preceding assessment. Current asthma was defined as having a parent-reported doctor-diagnosis of asthma at or before the assessment and either taking asthma medication and/or having an illness with wheezing that lasted for a week or more in the previous 12 months.

Results: At baseline, the infant and kindergarten cohorts comprised 5,107 and 4,983 children respectively, with 83% and 84% followed up after 4 years. At age 4-5 years, the cumulative incidence of ever diagnosed asthma was almost identical in the infant and kindergarten cohorts, as was the prevalence of current asthma. Almost one-third of children had been given a diagnosis of asthma by age 8-9 years.

Conclusions: The prevalence of current asthma increased with age, reaching a plateau at age 6-7 years. However, the cumulative incidence of ever being diagnosed with asthma continued to rise from birth to age 8-9 years. Analysis of Wave 4 data of the LSAC, will allow us to investigate further changes in these parameters over time.

Nutrition and Development

Katherine Thurber, Australian National University, National Centre for Epidemiology and Population Health

The Longitudinal Study of Indigenous Children (LSIC), run by the Australian Government Department of Families, Housing, Community Services and Indigenous Affairs (FaHCSIA), is the first nation-wide longitudinal study to address the physical, emotional, social, and spiritual wellbeing of Aboriginal and Torres Strait Islander children in Australia. Previous research has demonstrated a clear link between early childhood nutrition and physical, intellectual, and behavioural development. Few studies, however, have examined this relationship within the Aboriginal and Torres Strait Islander populations of Australia. The current study, through the use of the LSIC dataset, aims to elucidate the relationship between the nutrition and development of today's Aboriginal and Torres Strait Islander children.

During Waves 1 and 2 of LSIC, a 24-hour food recall was administered to all children in the older cohort (aged 3.5 – 4.5 years in Wave 1, and 4.5 – 5.5 years in Wave 2), providing an indication of the types, though not quantity, of foods consumed. Data were also collected about children's breastfeeding history, breakfast habits, and beverage consumption. In addition, researchers recorded the mothers' food, alcohol, and smoking habits during pregnancy, thus creating a detailed nutritional profile for each child. The LSIC also provides several measures of physical, intellectual, and behavioural development for both study cohorts. Anthropomorphic data were collected for each child in the older cohort, with greater reliability in more recent waves of the study. Where data proves to be reliable, changes in weight-for-height percentile, height-for-age percentile, and BMI can be used to determine trends in children's growth patterns. Subjective ratings of children's health state, together with history of health service usage, can provide further information about the children's physical health. The Renfrew Word Finding Vocabulary Test is administered to the older cohort to measure speech and language abilities, and the MacArthur Bates Communicative Development Inventory is used among the younger cohort (aged 0.5-1.5 years in Wave 1, and 1.5-2.5 years in Wave 2) to measure language development. Additionally, both cohorts complete some items adapted from the Parent's Evaluation of Developmental Status (PEDS), a screening inventory designed to measure developmental and behavioral irregularities. An analysis of the associations between early childhood nutrition and these developmental outcomes will provide information critical for the development of effective nutritional intervention programs to be established within these communities.

Childhood overweight and mental health problems: is there a dose-response relation?

Pauline Jansen, Fiona Mensah and Melissa Wake, Murdoch Children's Research Institute

Overweight and mental health problems are very common health problems in childhood and present a risk for future health and well-being. In adolescence and adulthood, there is a clear causal bi-directional relationship between the two conditions: overweight increases the odds for future psychosocial and psychiatric problems and evidence is also found for the opposite direction with mental health problems resulting in weight gain. The association is less explored among school-aged children, but there is some evidence supporting a bi-directional association becoming particularly apparent in the transition from childhood to adolescence. Longitudinal studies on this topic typically assessed behavioural problems of children at baseline and examined its influence on BMI several years later, or reversed for the opposite direction of the association. Such a design is particularly useful for identifying causality and direction of a relationship, but it remains largely unknown how difficulties in the period from baseline to follow up influenced the outcomes under study. Hypothetically, persistent problems –either behavioural or weight problems- might have more adverse effects on the outcome under study as compared to temporary problems.

The aim of this study is to examine the possible dose-response nature of the association between overweight and behavioural problems in childhood. This study takes advantage of four waves of data collection of the Longitudinal Study of Australian Children. As the fourth wave of data is not available at the moment, results cannot be presented yet. However, cross-sectional analyses of the first waves of the K-cohort of LSAC showed that overweight children tend to have more behavioural problems as measured with the Strengths and Difficulties Questionnaire (Total Difficulties), and Paediatric Quality of Life Inventory (psychosocial health summary). This association was statistically non-significant at the ages of 4/5 and 6/7, but became significant at 8/9 years. These results confirm previous studies suggesting that the weight-mental health relationship becomes particularly apparent in late childhood. In our study, these cross-sectional analyses will be expanded by examining the influence of persistence of weight and behavioural problems.

Do language problems predict emotional and behavioural disorders in children and adolescents: evidence from LSAC

Richard O'Kearney and Shaun Goh, Department of Psychology; The Australian National University

Communication and language problems in children and adolescents often co-occur with emotional and behavioural problems (commonly called internalising and externalising problems). It is not clear, however, what role language problems play in developing and maintaining emotional problems across childhood and into adolescence. During childhood, language becomes increasingly important in self-concept, emotion regulation and in enhancing adaptive interpersonal behaviours. There are a number of existing longitudinal studies and other sources, which provide relevant data to clarify this role with regard to emotional and behavioural disorders. The aim of this presentation is firstly to describe and present data from a systematic review of the evidence about the association of language problems and emotional and behavioural disorders in childhood and adolescence from prospective studies; and then to investigate the relationship of language abilities to emotional and behavioural outcomes in the Longitudinal Study of Australian children (LSAC). Preliminary analysis of LSAC indicates an increased risk of internalising outcomes at age 9 for children with language problems at school entry, which varies for boys and girls with girls at higher risk. The study will also examine whether the link between language problems and emotional and behavioural problems is different for boys and girls. Evidence from the 2 sources will enable hypotheses about how language difficulties interact with emotional and behavioural problems and whether there is a role for communication or language training in the prevention of internalising and externalising problems in childhood and adolescence.

Correspondence between communication impairment in early childhood and outcomes at school

Jane McCormack, Linda J. Harrison and Sharynne McLeod, Charles Sturt University and Lindy McAllister^,The University of Queensland

Confirmation of the high prevalence of communication impairment in early childhood for Australian children was demonstrated by McLeod and Harrison's analysis of Wave 1 LSAC data. They reported that 25.2% of parents were concerned about their 4 to 5-year-old child's ability to “talk and make speech sounds”. Subsequent analysis of Wave 2 data from the same group of children revealed that early communication impairment was associated with later difficulties with literacy and numeracy. This association between oral and written communication difficulties is supported by an international body of literature; however, less is known about associations with other developmental outcomes.

The present study extends the above analyses in three ways, by: (1) examining a further wave of longitudinal data; (2) examining the associations between communication impairment and a broad range of life activities including language and learning outcomes; and (3) drawing on multiple perspectives to provide a more holistic examination of outcomes.

The selection of outcome measures was guided by the International Classification of Functioning, Disability and Health – Children and Youth version (ICF-CY; World Health Organization, 2007), which provides a comprehensive list of activities in which children participate. Furthermore, measures were selected that had been completed by different participants, including parents, teachers, children, and trained interviewers. Using the Wave 3 sample of 4,329 children aged 7-9 years, comparisons were made between the 1041 (24%) children who had been identified at 4-5 years by their parents as having communication difficulties and the 3,288 (76%) children who had not been identified. For each measure, the proportion of children whose scores were within and outside the normal range was compared (using a cut-off of 1 standard deviation below/above the mean score). The unique and combined effects of communication impairment and four child and family covariates (sex, age, Indigenous status, and socioeconomic position) were also examined.

Children identified with communication impairment at 4-5 years performed significantly less well on all areas. Parents and teachers reported slower progression in reading, writing, and overall school achievement. Children reported more bullying, poorer peer relationships, and less enjoyment of school. This paper provides strong evidence that communication impairment in early childhood is associated with difficulties performing a range of life activities at school. Thus, there is a need to ensure ongoing intervention is provided, and to ensure this intervention addresses the full range of affected life activities.

Measuring the economic impact of speech impairment on academic achievement and social wellbeing: A study of the Longitudinal Study of Australian Children

Paula Cronin, Rebecca Reeve and Stephen Goodall, Centre for Health Economics Research and Evaluation, University of Technology

The prevalence of speech and language impairment in 4-5 yr olds could be as high as 20%. Increasing evidence suggests that there are both short and long term effects of childhood speech impairment. The impact of speech impairment may include loss of potential in occupational and educational domains and reduced quality of life.

The social and economic cost of communication impairment is significant to both individuals and society. High short- and medium-term costs are associated with children needing special education resources in school and repeating grades. High long-term costs to the individual (and society) are associated with children failing to finish school.

To date there is little if any studies that have addressed the economic impact of speech impairment in Australian Children. The first stage of this study is to investigate the impact of speech impairment on educational outcomes and wellbeing.

Three waves of the Longitudinal Study of Australian Children (LSAC) (Wave 1,2 and 3) (N~2340) was analysed using a random co-efficient model. The model examined the effects of speech impairment (measured by 4 measures: PPVT (Peabody Picture Vocabulary Test ) assessment, parents expressive language concern and receptive language concern, speech therapy ) on 5 outcomes (literacy, maths reasoning, approach the learning, school adjustment and social-emotional wellbeing).

Results indicate that children with below average PPVT (direct assessment of receptive language impairment) had significantly lower maths and literacy scores in Year 2 and 3 (- 7.49% SE 0.00646), (-7.42% SE 0.0113) respectively. Children entering speech therapy at an older age had lower levels of school adjustment (-0.07 SE 0.0474) and parents receptive language concern was negatively associated with children's self esteem at 6/7 and 8/9 yrs (social-emotional) (-8.34% SE 0.0130). Children with a positive approach to school (both across individuals and changes in an individual over time) had higher literacy, (14.8% SE 0.00485) numeracy (9.26 % SE 0.00916) , school adjustment (0.0316 SE 0.0159) and self esteem. (2.58% SE 0.00695).

This study demonstrates there is a link between early interventions for SLI and improved outcomes in literacy and numeracy. Since there are clear links between school achievements and productivity gains the economic impact from early interventions for SLI could be significant. Further research is warranted.

Fathers' involvement in young children's social and educational activities

Jennifer Baxter, Australian Institute of Family Studies

This paper uses the Longitudinal Study of Australian Children to explore fathers' involvement with children, in particular with their social and educational activities. These activities are eating an evening meal with children, talking with them about their day and helping them with homework. Fathers' involvement is measured with respect to one child in the family – the LSAC study child. These children were aged 2-3 years through to 8-9 years at Waves 2 and 3 of LSAC, when these particular items were introduced. The focus of this paper is on the characteristics of fathers that explain a greater degree of involvement with these different activities. This research finds that more paid work hours by fathers reduces involvement in these activities, but more paid work hours by mothers increases fathers' involvement. Fathers are also more likely to be involved in these activities if they have better mental health and if the child has a less reactive temperament. Fathers with higher levels of education spend more time talking with their children and helping them with homework. Fathers who have children living in another household, or with more resident children, spend less time on these activities. Other results about fathering in Australia will also be discussed.

What about Dad? How non-resident fathers influence child wellbeing

Nina Lucas, Parenting Research Centre, Bircan Erbas, La Trobe University and Jan Nicholson, Parenting Research Centre

Several decades of research demonstrate that, relative to those from intact families, children of separated parents have poorer outcomes on a range of measures. Explanations include poorer socioeconomic status of lone mother families, poor post-separation parenting, poor mental health of lone mothers, and exposure to parental conflict. Explanations so far have tended to focus on resident mothers, neglecting the potential importance of non-resident fathers. This study first aims to describe the population of children with non-resident fathers in Australia, relative to those with resident fathers. The second aim is to compare the mental health of children with resident and non-resident fathers, and determine whether differences can be explained by differences in socio-economic status, parenting, parent mental health, or parental conflict.

We use data from wave 3 (K cohort) of Growing Up in Australia: The Longitudinal Study of Australian Children. Child mental health was measured using teacher-reports on the Strengths and Difficulties Questionnaire. Other data were reported by mothers, resident fathers, and non-resident fathers.

Children with a non-resident father were more likely to be exposed to social disadvantage, poor parental mental health, and high levels of parental conflict than those with a resident father. However, non-resident fathers reported higher parental warmth and higher use of inductive reasoning than did resident fathers. Children with a non-resident father had 2.3 times the odds of a mental health problem than those with a resident father, but this difference was explained by differences in demographic characteristics, socioeconomic status, parenting, parent mental health and parental conflict. The strongest predictors of child mental health were fathers' parenting, mothers' and fathers' employment, and mothers' education.

The poor mental health of children with separated parents can be explained by differences in socioeconomic circumstances, parenting practices, parent mental health, and exposure to parental conflict. Fathers' parenting was a key driver of child wellbeing, even after adjustment for mothers' parenting. The findings suggest that policies should focus on improving the financial and psychological wellbeing, and parenting skills, of mothers and fathers post-separation.

Indigenous family relationships and post-separation parenting

Maggie Walter, Department of Sociology and Social Work, University of Tasmania and Belinda Hewitt, Institute for Social Science Research, University of Queensland

This paper explores the shape and practice of Indigenous family relationships as they relate to post-separation parenting. While we already know that Indigenous families are younger, poorer and more likely to be sole parent families than non-Indigenous families we know very little about how these families practice the complexities of post-parental separation parenting. Data from LSIC Wave 2 allows for the mapping of these dimensions of family life. In this study we investigate, for those children whose other parent lives in another household, the pattern of study children's interaction with non-resident parents and on-going financial support from that parent primarily as reflected by the payment or receipt of child support. Our results indicate that nearly 40 percent of our study children have a parent living elsewhere. Additionally, while not all such families are sole parent families, the data find that significant and systematic disadvantages accrued to study children who resided in a household with only one parent. Around three quarters of children with a parent living elsewhere, however, had at least some contact with their other parent with frequency of contact varying by the age of the child. Also, nearly 60 percent of primary parents whose study child had a parent living elsewhere received child support. The majority received their payments via the Child Support Agency while around 40 percent of parents had private arrangements for the transfer of child support monies. Finally, although whether or not receipt of child support provided major benefits for the household is not completely clear, there is some indication that primary parents who received child support had higher household income than those that did not.

Persistent inequality? A comparison of the impact of family background on children's outcomes in the UK and Australia

Jo Blanden, Department of Economics, University of Surrey and Centre for Economic Performance, UK and Ilan Katz (presenter) & Gerry Redmond, Social Policy Research Centre, University of New South Wales

This paper adds to the understanding of how family background, in particular parents' educational attainment, contributes to inequality in the next generation, and how different social contexts may impact on the inter-generational transmission of advantage and disadvantage.

We compare the cognitive and socio-emotional wellbeing of children from families with different levels of parental education (a measure of SES) over the first three waves of LSAC and the UK Millennium Cohort Study (MCS) to address the following issues:

How persistent are early differences in cognitive and social and emotional development?
How is wellbeing in both nations influenced by parental education?
Is there convergence or divergence in the outcomes of children from different backgrounds as they move into middle childhood? Is this similar in both nations?
What is the connection between parental education and persistence of different levels of wellbeing in the two countries?
Are the trajectories of initially good or poor performers influenced by their parents' education?

Three analyses of the two studies are presented:

A comparison of the mean differences in cognitive and socio-emotional wellbeing between children from different backgrounds at different ages in the two countries.
A comparison of the trajectories of children from different backgrounds with different levels of wellbeing at the first wave in each study
An examination of the role of parental education in the persistence of high and low cognitive and socio emotional wellbeing in the two countries.

Overall both these two countries show similar associations between parental educational attainment and children's outcomes in the cognitive, social and emotional domains. There are large and continuing association between children's developmental outcomes and parental levels of education for both countries. With regard to the differences between the two countries, our main finding is a consistent pattern of greater inequality driven by parental education in the UK than in Australia, with the biggest difference being that children in the UK with lowest parental education have relatively much lower levels of wellbeing in both domains compared to their compatriots than their equivalents in Australia.

The link between Indigenous culture and well-being: Qualitative evidence for Australian Aboriginals

Simon Colquhoun and Mike Dockery, Centre for Labour Market Research, Curtin Business School

Evidence from both the international and Australian literature suggests that the wellbeing of Indigenous people can be enhanced by maintaining their traditional culture. This paper uses qualitative data made available from the LSIC to explore this relationship in the context of Aboriginal and Torres Strait Islander Australians. Specifically, responses to two open-ended questions “What is it about Aboriginal or Torres Strait Islander culture that will help the study child to grow up strong?” and “Apart from health and happiness, what do you want for the study child?” are analysed using Leximancer, revealing a number of key themes from the responses. In relation to the question on culture, family is the dominant theme, while the other themes to emerge appear to relate to cultural identity, cultural presence, understanding of culture and a sense of belonging. In relation to the question on what parents want for their children, eight key themes emerged which we interpret as reflecting a balance of desires for success in the dominant culture (including education and success) and in their traditional culture (being strong, to have a close relationship with their family, to be whoever their children want to be). The responses to these two questions highlight that Indigenous parents see the importance of maintaining and learning about aspects of their culture for identity development, the positive experience of the traditional culture and significance of support from the community to which they belong in order to achieve the healthy development of their child. What is also clear is the impact of the dominant culture on this process, which, according to available research, can both influence and inhibit the healthy development of the Indigenous child.

Does parental social marital status matter to child well-being?

Lixia Qu and Ruth Weston, Australian Institute of Family Studies

The increasing proportion of couples living together without having married (here called “cohabitation”) represents one of the most significant changes over the last few decades in Australian family life. This trends has been accompanied by an increase in the proportion of babies are born outside marriage. In 2009, 34% of births registered in Australia were ex-nuptial compared with 12% in 1980. Most of these births were to mothers in cohabiting relationships. So far, limited attention has been given to possible implications for children born into such families, especially in Australia. The overseas studies suggest that children living with cohabiting parents tend to do less well developmentally and have access to limited economic resources compared with children with married parents.

This paper uses the data from three waves of the Longitudinal Study of Australian Children. The following issues are examined for three groups of children (children living in cohabiting two-biological-parent families, children living in married two-biological-parent families and children living in sole-mother families): (a) parental demographic characteristics and socio-economic status, (b) the quality of parental relationship and parenting practices, and (c) children's development in three domains (physical, social and emotional, and learning). Multivariate analysis will be applied to examine the extent to which differences in children's development (if any) could be explained by any systematic differences in family circumstances (e.g., socio-economic status). In addition, this paper examines children's development according to their parental cohabitation pathways (marriage, separation, and continuing cohabitation).

Children's Media Use and Time Choices

Leonie Rutherford, Faculty of Arts & Education, Deakin University and Jude Brown, School of Behavioural, Cognitive and Social Sciences, University of New England

Policy makers often wrestle with the question of how time spent in various activities affects children's development. It is the mandated responsibility of The Australian Communications and Media Authority (ACMA)—Australia's broadcast regulator—to monitor the extent of this media use and community attitudes towards it on a regular basis. Most ratings agencies panels, and such national surveys as ACMA's Families and Electronic Entertainment (1996) and Media and Communication in Australian Families (2007) study aspects of media choices by time of day. But even these large national surveys canvass only a relatively small number of children from each stage in the age range (0-17 years).

While the LSAC is an ‘omnibus' survey that was not designed specifically with media practices in mind, it contains a suite of instruments that can deliver precise measurements across a range of areas of developmental significance. In the case of media usage, the time diary, for example, offers more reliable measures of time investments, and where they occur during the day and week for individual children, than most large-scale surveys, as well as some information about the social context of usage. This allows researchers to map patterns of media use and other ‘lifestyle' choices longitudinally to see how they affect developmental outcomes.

Our paper offers a progressive story about the developmental stages of children's media use and lifestyles during a period in which digital platforms are widely seen to be displacing other, older, media practices. It analyses the ‘episodes' of media usage of Australian children, their durations, occurrences and timing, for each wave of each cohort, and examines the effects of demographic variables such as gender and SES. To offer a contemporary evaluation of the hypothesis that ‘old' media forms are displaced by ‘newer' forms, the paper also investigates patterns of ‘oldest' media use compared to new, testing whether use (or non-use) or high (versus low use) of particular media establishes patterns of behaviour that continue throughout a child's life course.

Parenting practices, children's media use and childhood obesity

Jude Brown, Helen Skouteris, Michael Bittman and Leonie Rutherford School of Behavioural, Cognitive and Social Sciences, University of New England

The impact of children's individual lifestyle behaviours on obesity is well known. As the family is the primary social force shaping children's lives, many of the risk factors for overweight and obesity in early childhood are likely to have roots within the family context. Based on their attitudes and beliefs, parents set and enforce rules and create a home environment that necessarily influences their children's participation in sedentary behaviours like television viewing or computer use, exercise and eating behaviours. Data from the first three waves of the K cohort (LSAC) data was used to investigate whether proximal and distal parental practices at Wave 1 and 2 (child aged 4/5 & 6/7) were associated with children's media use, and whether these lifestyle behaviours at Wave 2 (child aged 6/7) were associated with child weight status at Wave 3 (child aged 8/9). Results from the path model revealed that consistent parenting (Wave 1) was associated with other parenting practices and some of the children's media use and lifestyle behaviours at Wave 2. Further the model revealed that although children's television watching was associated with more snacking and less moderate to vigorous activity, only time spent in television viewing was significantly associated with child weight status at age 8/9 years. The analyses revealed a clear pathway linking consistent parenting with parental practices regarding television viewing, with the time children spent watching television and child weight status.

Is mother's time a public good with respect to child development?

Matthew Taylor, The National Centre for Social and Economic Modelling, University of Canberra

Many studies have found differences in educational attainment and earnings according to family size and birth order. One important determinant of these finding is likely to be differences in parental investments resulting from the sequential nature of childbirth. Price studied how much “quality” time children receive with parents and found that second born children receive less at the same age compared to firstborns.

A recent theoretical contribution by Gugl and Welling posits that older children benefit from a greater amount of time spent with their primary carer, first when they are an only child and later as the arrival of second child decreases the probability that the primary carer will enter the labour force. This theory is based upon an assumption that the primary carer's time is a “public good”, that the developmental advantage of the primary carer's time is not diminished by the presence of other children in the home (i.e. mother's time is not rival in consumption from the children's' perspective).

This paper presents descriptive evidence of differences in child outcomes according to family size and birth order for both LSAC cohorts. We then use differences in school starting age policies across each Australian state and territory that move the study child's older siblings into school in different years as a source of exogenous variation in the amount of time B cohort children spend as the only child in the home. Comparing the outcomes of B cohort children who share the home with siblings at a given age with those who do not, by virtue of school entry legislation, provides a robust test of our hypothesis that mothers' time is a public good. This research will make use of three waves of Time-Use Diary data and focus on a range of child outcomes measured in LSAC.

Analysis of the breadth, severity and stability of child health inequalities

Fiona Mensah, Jan Nicholson, Murdoch Children's Research Institute, Liz Headley, John B. Carlin, Department of Paediatrics, University of Melbourne, Donna Berthelsen, Centre for Learning Innovation, Queensland University of Technology and Melissa Wake, School of Public Health and Preventive Medicine, Monash University

Background: The impact of socio-economic inequalities on health and development is particularly marked when multiple outcomes are considered. This paper explores latent class analysis as a method of integrating multiple outcomes to examine a breadth of health inequalities.

Objectives: To investigate which groups of children can be identified, and whether similar groupings are replicable at different ages. To examine how far individual children maintain group membership over time and examine inequalities on the basis of these profiles.

Method: Data were from the Longitudinal Study of Australian Children, the birth cohort at ages 4-5 years, and the Kindergarten cohort at ages 4-5, 6-7 and 8-9 years (~5,000 children each). Measures included global health; special health care needs; asthma; sleep difficulties; injuries; body mass index; quality of life; behaviour; vocabulary; school readiness; and academic competence. Subgroups of children within each cohort at each age were identified using latent class analysis. Reproducibility of the groups was examined by comparing the Birth and Kindergarten cohorts.

Results: Similar groupings were identified within each of the cohorts at each age. In the Kindergarten cohort at age 4-5, 59.5% of the children were classified as the healthiest group, 33.4% had a moderate level of difficulties and 7.2% had the most severe level of difficulties. Continuity between children's group membership through ages 4 to 9 was strongly evident. Socio-economically disadvantaged children experienced severe difficulties most frequently, in the Kindergarten cohort at age 4-5, 2.8% of girls in the most advantaged families had a severe level of difficulties compared to 17.2% of boys in the most disadvantaged families.

Parental Health, Health risk Behaviours and Child Weight

Preety Srtivastava, Centre for Health Economics/ Department of Econometrics & Business Statistics, Monash University and Xiaohui Zhang (presenter), Department of Econometrics and Business Statistics, Monash University

Child obesity has now become a global epidemic in developed countries and a potential emerging public health issue in developing ones. The associated health problems and costs are raising concerns among health care professionals and policy makers worldwide. Obesity in children is among the most important risks to children's short-term as well as long-term health. It causes a wide range of serious complications and increases the risk of premature illness such as type 2 diabetes, cardiovascular diseases, and mortality.

Economic studies on child obesity are burgeoning and have mostly focused on the effect of maternal employment and on US and UK data. While the relationship between maternal employment and child obesity is now well established in the literature, not a single study has explicitly examined the impact of parental health and health risk behaviours on child obesity. Whether in matters of social development or physical health, parents are typically considered the primary agents of children's socialisation and well-being. Thus parents' lifestyle habits such as healthy eating and exercise are most likely to have an important impact on children's lifestyle, which in turn would affect their health.

The effectiveness of any health policy and education program to address child overweight requires a good understanding of the pathways leading to the problem. Empirical studies are crucial to identify and quantify these pathways. The broad objective of this study to examine causal pathways between child weight and parental health and health risk behaviours using the 3 waves of the Longitudinal Study of Australian Children (LSAC), a survey that follows the development of over 10,000 children and families from all parts of Australia. Besides the direct causal effect, parents' health behaviours and child weight may also be associated via unobserved characteristics such as parents' time preference. We therefore propose to use advanced micro econometric techniques in this study, including a fixed-effect model to examine the interrelationship across children's obesity, lifestyle and food intake; parents' health status and health risk behaviours; and household demographic and socioeconomic factors.

Mental health in ‘low-to-moderate risk' preterm, low birth weight and small for gestational age children at 4-5 years: The role of early maternal parenting

Elizabeth Westrupp, Fiona Mensah, Rebecca Giallo, Amanda Cooklin and Jan Nicholson, Murdoch Children's Research Institute, Parenting Research Centre, Clinical Epidemiology and Biostatistics Unit, Royal Children's Hospital, Department of Paediatrics, The University of Melbourne

Objectives: The majority of children born preterm, low birth weight or small for gestational age are born with low-to-moderate risk (LTM), yet most research focuses on the high-risk group. Little is known about whether children with LTM perinatal risk are at greater risk of mental health problems, or the role of early maternal parenting in determining these outcomes.

Methods: Longitudinal data were from a large nationally representative Australian cohort of 5,000 children, aged 0-1, 2-3 and 4-5 years of age. Participants were 354 children with LTM perinatal risk (born <37 and >32 weeks gestation, birth weight <2,500 and >1500 g, or <10th and >1st percentile for gestational age) and 2,461 children in the normal birth weight, term comparison group. Child mental health was measured using maternal-reported total difficulties on the Strengths and Difficulties Questionnaire (SDQ). Parenting irritability, warmth, self-efficacy, maternal separation anxiety and overprotective parenting were measured when children were age 0-1 and 2-3.

Results: Parents in the LTM perinatal risk group were more likely to experience parenting difficulties on 1 of 8 parenting measures (irritable parenting at age 0-1) when adjusting for socio-demographic differences (OR = 1.43; 95% CI = 1.05, 1.95, p<0.02). This group difference was no longer apparent by age 2-3. LTM perinatal risk group membership predicted mental health problems on the SDQ at age 4-5 in the unadjusted model but not when accounting for maternal-specific and socio-demographic differences.

Conclusions: Children with LTM perinatal risk did not significantly differ in their mental health from other children of similar social backgrounds. These findings support an environmental, rather than parenting, pathway to psychological risk in children born with LTM perinatal risk.

Economic cost of common childhood disorders

Nina Lucas, Jon Quach, Emma Sciberras, Elizabeth Westrupp, Jan M. Nicholson, Jordana Bayer, Louise Canterford, Daryl Efron, Lisa Gold, Harriet Hiscock, Fiona K. Mensah and Melissa Wake, Parenting Research Centre

Policy makers are increasingly being asked to set priorities for spending on child health interventions. To plan services efficiently, information is needed about the economic costs of common childhood disorders, and how these costs vary according to the severity and stability of these disorders over time. In this symposium of linked papers, we examine the costs associated with four common childhood disorders: perinatal risk, mental health problems, sleep problems, and attention deficits.

These projects use linked Medicare data to estimate the economic costs to government associated with common childhood disorders in the LSAC cohorts. Costs estimated are those spent on the Medicare Benefit Schedule (MBS) and Pharmaceutical Benefit Scheme (PBS) for LSAC children from birth to age 4 years, and from age 4-8 years.

The extent and socio-economic determinants of early Indigenous Australians' disadvantage in cognitive development

Richard P.C. Brown*, University of Queensland, Colin Green, Lancaster University and Prabha Prayaga, University of Queensland

Indigenous Australians are among the most disadvantaged groups in developed countries. This is reflected across a range of socio-economic and health outcomes. Recent research has demonstrated that Indigenous Australians suffer from significant educational disadvantage and moreover that this disadvantage is of a substantial magnitude at the time of entry into formal schooling. Educational attainment is a key predictor of later life socio-economic outcomes. This suggests a need to intervene in educational attainment, and cognitive development more generally, in the pre-school period. This paper adds to this research by using the Longitudinal Study of Indigenous Children (LSIC) to examine measures of the cognitive development of Indigenous children in Australia prior to school entry. These cognitive skills provide important building blocks for later educational attainment. In particular we seek to determine the extent and timing of pre-school disadvantage in cognitive development amongst Indigenous Australians. In turn, we examine the socio-economic determinants of this cognitive development. This is of particular importance as appropriate policy responses will vary markedly according to the sources of the disadvantage (health, parental involvement, negative life events etc). Alternative policy interventions will be assessed in the light of our findings.

Describing Indigenous communities covered in LSIC and with some preliminary exploration of data quality of community variables in LSIC data

Boyd Hunter and Matthew McKay, Centre for Aboriginal Economic Policy Research (CAEPR), Australian National University

LSIC provides a potentially useful source of information on Indigenous children, but it is important that users understand some basic aspects of the survey design and data quality. This paper attempts to provide some useful background by identifying some of the geographical nuances of the LSIC data with an initial focus on how study communities differ from other similar communities. The analysis makes extensive use of information on Indigenous areas from the 2006 census to describe LSIC communities, but also tests the internal consistency of the data with a local geographic dimension in the first wave of the study. The Intra-Correlation Coefficient (ICC) will be used to document the correlation of response for ‘community' LSIC variables within Indigenous Area geography. ICCs can be estimated for data derived from the following questions: Is this a good community or neighbourhood for little kids? Are there good places for kids to play in this community or neighbourhood?; is this a noisy neighbourhood?; and finally how safe would you say this community or neighbourhood is? This will not be perfect validation of the data but will allow the user to be more confident that the data is reasonably consistent with the local geography. Some preliminary exploration of the association between community level data from the census and LSIC and some rudimentary child outcomes will be attempted.

Social inclusion and participation in community activities for children with disability

Amanda Knight and Leo Bild, Department of Families, Housing, Communities and Indigenous Affairs (FaHCSIA)

Literature suggests families with a child that has a disability are more likely to experience increased risk and vulnerability than other families. One such risk is social isolation of both child and parent. Families with a child who has a disability often need to build a level of resilience to reap the benefits of social inclusion. The ‘benefits of inclusion have been identified as providing children with disabilities the opportunities for social, educational and language development'.

There is a breadth of research around childhood disabilities and engagement within the education environment but much less on the social community context especially for children with severe disability. For those children social engagement can help increase psychological and physical wellbeing, which contributes to positive relationships and social interactions. Social engagement is also an important protective factor for families because it develops social competency and self-esteem. A further component of this protective factor is the ability to access government support services. Social connectedness of children at this early age is an important base to build their interaction skills so that they will develop as they enter the education system and begin to interact with peer.

This study investigates the relationship between children with disability and those without using B and K cohorts multiple waves. To investigate this area we identified indicators of disability, school attendance, service access and unmet need along with community activity engagement.

Identifying problem areas where children with disabilities have low levels of community engagement will potentially help policy makers grappling with strategies for social inclusion. This paper will highlight differences between the experiences of children with disability and those without. The policy implications will become clear and will be directed from the children's experiences and those of their parents.

A tyranny of distance or disadvantage? Children living in rural and regional Australia

Ben Edwards and Jennifer Baxter, Australian Institute of Family Studies

There is increasing policy interest in the issues experienced by families in rural and regional Australia. However, little research has been undertaken into the circumstances experienced by children living in these areas. In this paper we compare children's receptive vocabulary and non-verbal reasoning by whether children are living in inner and outer regional areas and major cities. Given that many regional areas are also disadvantaged, we test whether differences in children's cognitive development are due to living in areas of relatively high unemployment rather than due to living in regional areas. This paper uses the first three waves of data from Growing Up In Australia: The Longitudinal Study of Australian Children augmented by data linkage with area based unemployment rates and census data. Findings from random effects regression models suggest that, at least for children's cognitive outcomes, there are differences by region and by the unemployment rates of the area. We extend our analyses to try to explain whether these differences are due to demographic characteristics, neighbourhood and social contexts and parent and family variables and find that once these factors are taken into account, children in the regions still have lower cognitive outcomes than in major cities but there are no statistically significant differences by the level of unemployment in the area. Finally, we take into account the educational aspirations of parents and the educational levels of parents of young children in the area to explore the role that parental and normative expectations play in shaping children's cognitive development. Findings from the current study have the potential to inform several policy agendas including enhancing the lives children living in the regions and informing aspects of the social inclusion agenda particularly area based initiatives.

Data workshop

Data Workshops were held on the day following the conference. The focus of the training was to assist users of the data, those considering becoming users, or those who are interested in learning more about LSAC data, to gain confidence in understanding and navigating the LSAC datasets. The training covered a range of topics designed to give a comprehensive overview of the conduct of the study, its datasets and supporting documentation.

To register your interest or find out more, contact the LSAC Team at AIFS.