1. Parenting and families in Australia
- 1.1 Introduction
- 1.2 Report overview
- 1.3 Section summary
- 1.4 Background to LSAC
- 1.5 Brief review of the literature
- 1.6 Description of method
- 1.7 Methods of data collection
- 1.8 Description of the LSAC sample
- 1.9 Discussion
1.1 Introduction
In opening their monograph on the wellbeing of Australia’s children, Richardson and Prior (2005) reflect:
Whatever the social construction of childhood and the allocation of care responsibilities, contemporary research makes it quite clear that the experience of childhood has a powerful and lasting impact on adult outcomes. We all have a deep interest in the quality of contemporary childhood, for at least the reason that from there will come the adults who comprise our future society. We also emphasise that the experience of childhood is important for its own sake. Children are a substantial part of our society, and their ill-being or wellbeing matters as much or more than does the wellbeing of adults: children are beings as well as
becomings. (p. 4)
While it is still the case that, throughout time, families have remained uniquely placed to bring children into the world and raise them, the contexts in which families do this are substantially different from those of 50 or even 20 years ago.
Australian families are smaller. The specialisation of households, with mothers caring for children and fathers ‘winning the bread’, has changed with the entry of increasing numbers of women into paid work. Labour markets have changed too, with substantial demands for greater levels of skill and the rise of part-time and casual work. Marriages, when they form, do so later and significant numbers of these end with periods of lone parentage followed by family re-formation. As the two-parent working household has emerged, so too has the demand for child care outside of the home. Previously unpaid because it was provided by women at home, the real cost of this care has now been transferred to families and governments as parents seek alternative care to enable participation in the labour force. This has given rise to concerns about the quality of care as it relates to child development. Amid this, the population is both living longer and the Australian demography ageing as the large proportion of baby boomers prepares to retire. Concurrently, the government is encouraging their retention in the workforce in the face of workforce skill shortages. Clearly there are competing demands outside of the family that now shape how families manage the task of raising children.
The impact of these changes on the capacities of growing children is uncertain. Some of this uncertainty stems from the absence of evidence of these effects upon families and on the outcomes for child development—a situation which has begun to change with the advent of ‘Growing Up in Australia’—the Longitudinal Study of Australian Children (LSAC).
LSAC offers the first large-scale opportunity to study the development of families and children in this period of unprecedented change. This report uses the data from the first wave of data collection of LSAC to examine how parents are faring in the tasks of parenting their children and how this relates to their day-to-day circumstances at home, with their partners, with family and friends, and in the local community.
1.2 Report overview
This report presents findings on parenting and family data from Wave 1 of LSAC, describes parenting practices and styles in different family contexts and investigates the level and type of support Australian parents receive in raising their children. The study of parenting touches on several of the key research questions that underpin the development of the LSAC methods and measures (Sanson et al. 2002). Primary among these are:
- What are the impacts of family relationships, composition and dynamics on child outcomes, and how do these change over time?
- What can be detected of the impacts and influences of fathers on their children?
- Do beliefs and expectations of children impact on child outcomes, and how do these change over time?
- How important are family and child social connections to child outcomes? How do these connections change over time and according to the child’s age? Does their importance vary across childhood?
While the content of this report is not designed to comprehensively address each of these questions, the report’s general objectives inform selected aspects of the questions. The report describes the relationships between factors that influence carers’ feelings and perceptions about the way they parent their children, about the way their families function and about the supports they use as parents. These in turn may affect developmental outcomes for children and the relationship of parenting practices to these outcomes as summarised in the LSAC Outcome Index.1 Five specific topic areas are explored in relation to all children in both cohorts, as well as in relation to children in couple and lone-parent families:
- parenting styles and family functioning
- factors influencing parents’ feelings and perceptions about the way they parent their children
- the roles and contributions of parents who do not live with their children
- feelings of stress and sources of social support for parents
- the relationship of parenting practices to child outcomes.
This report begins with a literature review that draws heavily on the previous publications of the authors and their colleagues. In particular, Nicholson et al. (2004), Nicholson et al. (2006b), Sanders, Gooley and Nicholson (2000), Sanders, Nicholson and Floyd (1997), Sanson et al. (2002), and Spence (1996). Interested readers are referred to these sources for more detail.
Sections 2 to 6 are constructed around the five specific topic areas. Readers are referred to the Appendix section of this report for additional information in the form of tables and figures that relate to specific topic areas. For instance, additional tables and figures for Section 2 can be found in Appendix B. A technical appendix (Appendix A) explains in detail analyses undertaken for each section.
1.3 Section summary
This section provides an overview of the conceptual model and research literature underpinning the measurement of parenting and family functioning for LSAC. The methods of sample selection, recruitment and data collection are summarised and a brief overview is presented of the characteristics of the infant and child participants, their primary carers (P1s), secondary carers (P2s) and the structure of their families.
1.4 Background to LSAC
LSAC is the first comprehensive national study to examine the lives of Australian children at regular intervals across infancy and early and middle childhood. LSAC aims to contribute to an understanding of children’s development in Australia’s current social, economic and cultural environment. The study is funded as part of the Australian Government’s Stronger Families and Communities Strategy by the Australian Government Department of Families, Housing, Community Services and Indigenous Affairs (FaHCSIA).
LSAC was announced in the 2000–01 Budget and work commenced on its development shortly thereafter. Sample recruitment and Wave 1 data collection were undertaken in 2004 and the data were released for analysis and research in May 2005. The study is being undertaken in partnership with the Australian Institute of Family Studies. Study design and implementation are supported by a nation-wide consortium of expert advisors (Sanson et al. 2002).
The broad intent of LSAC is to assist governments to develop effective policies on early childhood issues, particularly on early intervention and prevention strategies in the areas of health, parenting, family relationships, early childhood education, child care and family support.
1.5 Brief review of the literature
The conceptual framework underpinning the study design and measurement selection for LSAC is a pathways socio-ecological model of children’s health and development (Nicholson et al. 2006b; Sanson et al. 2002). Socio-ecological models acknowledge that the health of an individual arises from the complex interactions that occur between the individual and the environments in which he or she lives. As shown in Figure 1, the child’s health and development occur within the context of the family, school and community environments, each of which are influenced by broader socioeconomic, structural, cultural and political factors.
Figure 1: Socio-ecological contexts shaping children’s development
Source: Bronfenbrenner (1979).
This model has been further elaborated by social epidemiologists to elucidate the interacting pathways of influence between various contexts, and how these in turn interact with the inherited risks and resiliencies of the individual to determine life course health outcomes. The socio-ecological model shown in Figure 2 depicts the direct and indirect pathways of influence from more distal to proximal settings, highlighting the central role of parents, family and peers in influencing health and development, especially in early childhood. The family context, which includes factors such as parent–child attachment, parenting practices, couple relationships, family conflict, and parental risk and protective factors (including parental substance use, mental health and health behaviours) has been widely recognised as having a significant direct influence on children’s health, as well as moderating the influence of the more distal or macro social determinants. The model also identifies the potential for key institutions such as child care settings and schools to influence health and development.
Figure 2: Pathways socio-ecological model of health and development
Source: Modified by Nicholson et al. (2004) from Lynch (2000).
For simplicity, feedback processes have not been represented in the pathways model. However, it is recognised that these occur at all levels. Individuals interact with their environments and influence the circumstances to which they are exposed. For example, the influence of children’s antisocial behaviours on their relationships with parents and peers contribute to and maintain adverse interactions and add to the risks for poor psychosocial adjustment (Sanders, Gooley & Nicholson 2000). Conversely, some circumstances (for example, harmonious family relationships, social support), individual behaviours (for example, parental engagement with services and schools), and individual characteristics (for example, social skills) can have a protective effect that moderates the impact of other adverse circumstances and disrupts the pathways to poor outcomes.
The following literature review highlights the key measurement areas included in LSAC relevant to this report on parenting and family functioning. As such, it is selective and sets out the broad research paradigms on parenting, relating these to the specific topic areas. Particular attention is paid to relationships between, and links among, constructs of parenting and family functioning (Sanson et al. 2002).
Research approaches to the study of parenting effects
There is increasing evidence that parenting behaviours are instrumental in achieving many child behavioural and developmental outcomes (Collins et al. 2000). Broad research paradigms that support this assertion have included longitudinal studies, genetically informative research designs, and ecological designs that test multilevel effects of contexts outside of the family. In addition to these broad approaches, there has also been a long tradition of experimental studies based on observational techniques, as well as increasing examples of real world implementations of programs that seek to change the repertoire of parenting behaviours.
Quite powerful effects of parenting have been observed in longitudinal studies of child outcomes, where these effects can be differentially observed in families of varying composition and type (Lipman et al. 1998; Steinberg et al. 1994). In addition to longitudinal studies, research efforts that examine the nature and effects of parenting on children have widened considerably in the past decade. Formerly, the prevailing view in the ‘nature–nurture’ debate placed the basic mechanisms that linked parenting behaviours to child outcomes almost entirely in the domain of nurture—that is, as effects in the social environment of the child, and specifically in the behaviour of the parent(s). Now, there is a vigorous research effort using genetically informative research designs—studies and methods that include all or most family members, genetically related siblings (including twin studies) and step and adoptive children in which elements of biology and environment have been manipulated in ‘natural experiments’ or through the careful combination of specific observations to better disentangle effects (Plomin & Rutter 1998; Reise et al. 2000; Rowe 1994; Rutter et al. 2001). In addition to studies that seek to examine biological mechanisms that link parenting to child outcomes, researchers have also examined the way that other contexts mediate or moderate parenting and the effects of parenting. These contexts include those of the immediate neighbourhood (Brooks-Gunn, Duncan & Aber 1997; Brooks-Gunn et al. 1993; Leventhal & Brooks-Gunn 2000, 2004) and peer-group influences (Dishion et al. 1991; Mounts & Steinberg 1995).
Much of this research would suggest that the child is the passive object of parenting and that the flow of effect is from the parent to the child. However, studies of temperament and parenting (for example, Collins et al. 2000; Sanson, Smart & Hemphill 2002) show a wider range of bidirectional and interactive effects between characteristics of the child and parenting behaviour. For example, parents respond differently to children with different temperament characteristics, and children with different temperament profiles respond differently to the same parenting practices (Putnam, Sanson & Rothbart 2002). Quite apart from parenting behaviour being close to the child, being fairly regular and occurring over extended periods of time, parenting behaviour entails reciprocal actions between the parent and child (Bronfenbrenner 1979). It is important to remember that parenting is not one thing—it is multifaceted or a ‘package’ of effects (Parke 2002). Parenting influences the child’s development of: emotional regulation, exploratory behaviour, communication, self-direction, intellectual flexibility, introspection, self-efficacy in meeting life’s challenges, moral values, expectations, and motivation. In all likelihood, some of these effects undoubtedly flow to the parent as well.
Whether findings on mechanisms that link parenting to child outcomes are from longitudinal samples of children followed over time, from genetically informative designs, or from data that are contextually enriched, they suggest a set of causal mechanisms that are not only within families (both biological and environmental mechanisms), but in the wider community context as well. While the causal pathways including many of these effects are yet to be fully understood, the current evidence has been compelling enough for many agencies to implement programs that seek to support and develop better parenting practices or directly intervene to change the repertoire of parenting behaviour (Spoth, Cleve & Chin 1998; Zubrick et al. 2005). Government agencies charged with the responsibility of developing policies, as well as funding and delivering services, are also interested in assessing the various impacts of these on families and their ability to parent their children. It is in this context that the findings from LSAC have a particular relevance as they are the first comprehensive, national view in Australia of how families, and parents particularly, are meeting the challenges of raising a family.
Parenting
Previous research shows strong associations between parenting quality and child outcomes. Specifically, the lack of a warm, positive relationship with parents, insecure attachment, harsh, inflexible, rigid or inconsistent discipline practices, and inadequate supervision of and involvement with children, increase the risk that children will develop major behavioural and emotional problems. These include conduct problems, substance abuse, antisocial behaviour, and participation in delinquent activities (Coie 1996; Loeber & Farrington 1998; Patterson 1982).
However, adding complexity to this field of inquiry, what is regarded as ‘optimal’ or ‘quality’ parenting, varies with the age and competencies of the child. Parenting behaviours that are appropriate and effective with younger children may not be so with older children. As children become more developmentally competent in terms of self-regulation, communication, comprehension, introspection and autonomy, so too must parenting behaviours change to best meet the needs and extend the competencies of the child.
For the parents of children in the early years of life, three dimensions of parenting emerge from past research as having a significant impact on children’s subsequent health and development. These are parental warmth, hostile parenting and consistency.
Parental warmth refers to interactions between the parent and child that are characterised by affectionate behaviours, a high degree of positive regard, expression of enjoyment of the child’s company, interest and involvement in the child’s interests and activities, responsiveness to the child’s moods and feelings, and positive expressions of approval and support (Pettit & Bates 1989; Rothbaum & Weisz 1994). Warm, affectionate and responsive parenting has been consistently related to positive developmental outcomes for children (Pettit & Bates 1989; Rothbaum & Weisz 1994; Teti & Candelaria 2002; Zubrick et al. 1995), with good predictive power over periods of up to 10 years. For example, lack of parental warmth during infancy predicts subsequent child aggressive behaviour during preschool (Pettit & Bates 1989) and at age 10 (Bradley, Caldwell & Rock 1988). Low parental warmth during the preschool years is associated with a three-fold increased risk of problems such as conduct disorder, depression, substance use and other health risk behaviours by age 15 (Fergusson, Horwood & Lynskey 1994). Low parental warmth occurs more frequently in families characterised by adversity and conflict, lone-parent families and parents with depression (Patterson & Forgatch 1990; Webster-Stratton 1990).
However, parental warmth alone does not constitute developmentally effective parenting. The ways in which parents manage challenging or problematic child behaviours is also important. There has been extensive documentation of the types of discipline strategies that are associated with poor outcomes for children, notably in the development of disruptive, aggressive or conduct-disordered behaviours. Clear causal pathways to these outcomes have been established for parenting that is angry, irritable, physically punitive and marked by elevated levels of criticism, negativity and emotional reactivity. These hostile, angry or coercive patterns of parenting have been linked to the subsequent development of conduct-related behaviour problems, and to the maintenance and exacerbation of these problems over time (Kazdin 1987; Loeber 1990; Patterson 1982; Patterson, De Baryshe & Ramsey 1989; Snyder & Patterson 1995; Teti & Candelaria 2002).
The third dimension of parenting that becomes increasingly important after infancy is parental consistency.
This reflects the extent to which parents are firm and consistent in their interactions with children. It includes setting clear, developmentally appropriate boundaries and expectations for children’s behaviours and following through on stated intentions. Consistency is necessary for establishing child prosocial behaviours and, in combination with high warmth and low hostility, is regarded as providing an optimal foundation for healthy child development (Baumrind 1973; Patterson 1982; Rothbaum & Weisz 1994). In a large Australian population study, the combination of inconsistent and hostile parenting was found to be associated with elevated rates of child and adolescent delinquency, attention problems, social problems and somatic complaints (Silburn et al. 1996).
Additionally, parents’ attitudes and beliefs about their competence as parents are closely related to parenting quality. The extent to which parents perceive themselves as competent, being as good as or better than other parents, is highly correlated with parenting behaviours (warmth, hostility and consistency), parent psychosocial wellbeing, family conflict and children’s outcomes (Sanders et al. 1999; Silburn et al. 1996). An advantage of assessing parenting self-efficacy as a determinant of children’s outcomes is that it can be measured in a relatively simple and consistent way for parents of children of all ages—thereby providing one of the few measures of parenting that does not require developmental changes when used in a longitudinal study design (Nicholson et al. 2006b). For these reasons, a brief measure of parenting self-efficacy was included in LSAC.
Parent wellbeing and family functioning
Parent–child interactions occur within the context of the family. Parenting quality is influenced by a range of characteristics of individual family members (most notably, those of the primary carer), the social interactions between family members, the internal and external resources and supports available to the family, and sources of stress. These factors have been termed ‘disruptors’ (Patterson, De Baryshe & Ramsey 1989) as they ‘disrupt’ the parenting process, in either a positive or negative way which, in turn, impacts on children’s health and development (Sanders, Gooley & Nicholson 2000). For example, low parental warmth, high hostility and inconsistent parenting have been found to occur more frequently in families characterised by parental psychopathology (particularly maternal depression), high family conflict, socioeconomic disadvantage (for example, low income, young maternal age, lone-parent status), lack of social support, inadequate use of preventive health services, high levels of exposure to adverse life events, and family transitions (for example, parental separation, frequent relocation) (Patterson, De Baryshe & Ramsey 1989; Sanders, Gooley & Nicholson 2000). Such families have been identified as placing children at increased risk for a range of poor socio-emotional, behavioural, physical and learning outcomes (Sanders, Gooley & Nicholson 2000; Spence 1996).
LSAC was designed to assess children’s exposures to risk and protective factors within each of these key domains as described below.
Individual parent wellbeing
Parents of children with socio-emotional problems are more likely to have significant mental health problems themselves than parents of non-problem children (Sanders, Gooley & Nicholson 2000). In particular, depression, anxiety or antisocial behaviour in mothers, and antisocial behaviour and substance use in fathers are associated with increased rates of depression, anxiety and conduct problems in children (Gelfand & Teti 1990; Reeves et al. 1996). Parental psychosocial disorders may impact on children directly through the genetic transmission of personal risk characteristics, by the modelling of inappropriate behaviours, or via disruptions to parent–child interactions (Patterson, De Baryshe & Ramsey 1989; Spence 1996). For example, parental depression is associated with higher rates of parent–child negativity and physical punishment, lack of warm, affectionate behaviours, and inconsistent parenting (Gelfand & Teti 1990).
In a population sample of children, a relatively small proportion of parents at any given assessment point will display clinically significant disorders such as depression or antisocial personality disorders. Consequently, it is more appropriate in population studies to assess for the presence of psychosocial symptoms rather than clinical disorders (Nicholson et al. 2006b). This approach was adopted in LSAC. It has the advantage of increasing the proportions of parents likely to be identified as having psychosocial problems, but has the disadvantage of being more weakly associated with children’s outcomes than clinical measures.
To supplement the measurement of psychosocial problems, two other aspects of parental affective wellbeing were included in LSAC. These were parents’ perceptions of the extent to which they were currently experiencing life difficulties and how well they felt they were coping with life in general. Inclusion of these types of broad subjective measures of stress and coping have been found in other Australian population samples to be associated with parent physical and mental health and family conflict (Silburn et al. 1996).
Couple relationship quality
For children in the early years of life who live in two-parent families, the quality of the family environment largely reflects the quality of the relationship between parents. For example, as is reported later in this section, 40 per cent of all infant cohort members were the only child in the household. For this reason, LSAC has focused on assessing the quality of couple relationships rather than employing measures of broader family functioning (Nicholson et al. 2006b).
Conflict between parents has long been recognised as a significant predictor of children’s socio-emotional wellbeing (Loeber & Stouthamer-Loeber 1986; Sanders, Nicholson & Floyd 1997; Spence 1996). Children exposed to conflict that is more frequent, intense and overt show higher rates of aggression, anxiety, depression and physical health problems. Conflict may impact directly on children through the modelling of interpersonal aggression and the physiological effects of exposure to threatening situations. However, inter parental conflict is also associated with disrupted parenting—in particular, lack of parental responsiveness, and increased parental irritability and negativity (Sanders, Nicholson & Floyd 1997).
Additionally, conflict that is specifically related to parenting and children’s behaviour has been found to be associated with socio-emotional problems in children. Independent of the effects of marital distress, higher levels of conflict over parenting have been found to be associated with clinically significant levels of behavioural disorders and anxiety in early and middle childhood (Dadds & Powell 1991). In couple step families, families presenting for clinical interventions were found to have elevated rates of parenting conflict despite an absence of couple relationship distress (Nicholson & Sanders 1999).
Conversely, couple relationships that are characterised by a high degree of warmth, happiness and mutual support are believed to promote child wellbeing and act as a buffer against other adverse risk exposures (Sanders, Gooley & Nicholson 2000). Positive couple relationships consistently predict warm parenting behaviours and are associated concurrently and prospectively with reduced risks for behavioural problems in preschool and adolescence (Howes & Markman 1989; Miller et al. 1993).
Three aspects of the couple relationship were assessed in LSAC: overall relationship happiness, level of conflict (argumentativeness) and partner support for parenting. A Queensland population survey of parents found that adverse scores on each of these three aspects of the couple relationship were significantly associated with having a child with significant behavioural or emotional problems (Sanders et al. 1999).
External stressors and supports
High rates of exposure to adverse life events are associated with poorer outcomes for children, including anxiety and conduct problems. An accumulation of parental exposures to adverse events—such as a death or illness in the family, job loss, victimisation—is associated with increased parental irritability, and a deterioration in parenting practices (Wahler & Dumas 1987). Adverse life events have been found to prospectively predict childhood infections, accidents, and admissions to hospital, under-utilisation of preventive preschool services, and conduct and poor educational outcomes in later childhood (Fergusson, Horwood & Lawton 1990; Fergusson et al. 1989).
Social support for parents acts as a protective factor against a range of poor outcomes for adults and children, and may buffer the effects of other risk factors (Sanders, Gooley & Nicholson 2000). For example, the children of depressed mothers who receive a high degree of social support show fewer behavioural problems than children of depressed, unsupported mothers. In contrast, parental isolation and lack of support is associated with parental depression, irritable and ineffective parenting (Billings & Moos 1983; Patterson & Forgatch 1990).
Socioeconomic, structural and demographic characteristics of families
One of the family factors that has been most consistently associated with poor physical health, socio-emotional health and learning outcomes for children is low family socioeconomic status (Fergusson, Horwood & Lawton 1990; Nicholson et al. 2004; Spence 1996). Low socioeconomic status is typically defined on the basis of a range of parent characteristics such as poor parental education, low income, young age at first parenthood, unemployment or low occupational prestige and minority ethnic status. Socioeconomic factors are inextricably linked to family structural factors. For example, when compared to Australian children living in two-parent families, children in lone-parent families are two to five times more likely to be in the low income quintile, to live in a household where the main source of income is from government benefits and to live in overcrowded conditions (Nicholson et al. 2004).
When considered in combination with exposure to adverse life events, family socioeconomic status across childhood has been found to account for a significant proportion of adverse outcomes in late childhood, middle adolescence and early adulthood. These poor outcomes span physical health (illness, infections, injuries, cardiovascular health, oral health), mental health (conduct problems, delinquency, depression, suicidal ideation), poor academic achievement (early school leaving, lack of school qualifications), risky health behaviours (substance use, unsafe sexual activity), poor employment history and adult relationship difficulties (Fergusson, Horwood & Lawton 1990; Fergusson, Horwood & Lynskey 1994; Poulton et al. 2002).
Socioeconomic adversity is associated with a range of other risk factors including increased parental psychopathology, lack of education and financial resources, low social support, inadequate housing and residence in underresourced and unsafe neighbourhoods (Spence 1996). The impact of socioeconomic disadvantage and lone parenthood on children’s outcomes is thought to be mediated by disruptions to parenting and family functioning, with positive parental mental health, quality parenting and positive partner and external supports protecting against negative impacts for children (Spence 1996).
Parents living elsewhere
Finally, it is important to consider the impact on children of parents who reside outside the child’s current household. In Australia, by ages 12 to 18 years, between 21 per cent and 27 per cent of all children have lived some portion of their childhood years in a lone-parent family (de Vaus & Gray 2003). In the majority of these families (around 93 per cent), children have a biological parent living elsewhere (de Vaus 2004), who may or may not provide care for the child. Past research has shown that a substantial proportion of non-resident parents have little or no contact with their children and that, over time, the frequency of contact declines for many non-resident parents and their children (Smyth & Fehlberg 2002). Frequency of contact also varies with the child’s age at the time of parental separation and with the child’s current age, and is greater for non-resident parents who live closer to their child’s home. Post-divorce parenting arrangements are also closely related to the post-divorce financial arrangements and the level of conflict that occurs between the separated parents (Smyth, Sheehan & Fehlberg 2001). Demographic differences abound in these arrangements.
Relatively little is known about the impact non-resident parents have on their children’s health and development, or what determines effective family functioning when parents live apart from one another. LSAC represents a unique opportunity to examine these associations within the context of a population-based sample, for which a wide range of other measures of parenting and family functioning are also being collected (Nicholson et al. 2006a).
1.6 Description of method
Sample selection and recruitment
With facilitation by FaHCSIA, the Health Insurance Commission agreed that the Medicare database, the most comprehensive database of Australia’s population, could be used as a sampling frame for LSAC. Children in the scope of the survey were those infants aged 3 to 12 months and children aged 4 years 3 months to 5 years at the time of sample selection. A target sample of 10,000 was sought, equally divided between these two cohorts.
A two-stage clustered design, based on postcodes, was chosen to permit community-level effects to be measured and analysed, and to allow cost effective face-to-face interviewing. Every effort was made to ensure that the sample chosen would be as representative as possible of Australia’s infants and 4 to 5 year olds. The first stage of sampling entailed selecting postcodes and the second stage sampled children within these, allowing analysis of children within communities. Children in both cohorts were selected from the same 311 postcodes. An average of 40 children per postcode in the larger states and 20 children per postcode in the smaller states and territories were selected for the study.
Stratification was used to ensure proportional geographic representation for states/territories and capital city statistical division/rest of state areas. Postcodes were randomly selected with probability proportional to size selection where possible, and with equal probability for small population postcodes. Children were randomly selected with approximately equal chance of selection for each child (about one in 25). Due to excessive data collection costs, some remote postcodes were excluded from the design, and the population estimates have been adjusted accordingly.
The selection of children and corresponding fieldwork occurred in four phases. This was done to enable sample selection of children born across all months of the calendar year, to attempt to reduce the age range of children at interview, and also because some of the target infant population had not been born at the time of the first phase selection.
The final Wave 1 sample represents 53 per cent of all families who were sent a letter by the Health Insurance Commission. After excluding non-contact the achieved response rate was 64 per cent for the infants and 57 per cent for the child cohort. Children with mothers or fathers who have completed Year 12 are a little overrepresented in the final sample. Infants with no siblings are underrepresented (by 3 percentage points), while 4 to 5 year olds in couple families are overrepresented and those in lone-parent families underrepresented (by 4 percentage points each). Broadly the LSAC sample is representative of the Australian population with no large differences from Australian Bureau of Statistics (ABS) census data on most characteristics. Comprehensive details on the design and sample are available elsewhere (Soloff et al. 2006).
To compensate for the differences between the final LSAC sample and the national population, weights were developed (see Soloff et al. 2006 for details). Analyses in this theme report entail the use of sample weights, and multivariable analyses are fully adjusted for the sample design (see Appendix A for more detail).
1.7 Methods of data collection
Study informants for Wave 1 include:
- the primary carer (P1)
- secondary carer, other resident parent or step-parent (P2)
- child care providers (formal or informal)
- pre-school or school teachers
- the child her/himself (physical markers and direct assessment tasks)
- some interviewer observation of the external environment.
The primary respondent is the child’s ‘primary carer’ (P1), defined as the person who knows most about the child and their birth, history and current routines. Parents were asked to nominate who was the primary carer for this purpose, and typically this was the child’s biological mother. The ‘secondary carer’ (P2) refers to anyone else resident in the household with a parental relationship to the child or the partner of the primary carer. It should be noted that the terms ‘primary carer’ and ‘secondary carer’ are used in this report to identify the carers who completed different aspects of data collection as described below. The terms should not be interpreted as implying that the primary carer has a greater attachment to the child than the secondary carer.
For the first wave of the study, the base design data collection entailed an interviewer spending one to two hours in the home to:
- obtain detailed information about the child, plus some information on the parent, from the primary carer—this covered the key areas of health, family functioning, parenting, education, child care and social support
- obtain sociodemographic information on the family (such as household structure and parental labour force status, educational attainment and income)—this was obtained from either the primary carer or the secondary carer
- leave behind self-complete modules for both primary carer and secondary carer covering other aspects of family functioning, health and support, which took about 20 minutes to complete—where time permitted, these were completed while the interviewer was in the home
- undertake physical measurement of the child (such as height, weight, girth and head circumference)
- administer the ‘Who am I?’ school readiness test and Peabody Picture Vocabulary Test of receptive language to the 4 to 5 year-old children
- obtain consent to contact any child care provider or teacher, plus contact details for the parents so they can be located for future waves.
For full information about the interviews and content, see Soloff et al. (2003).
Measurement of parenting and family functioning
The reliable and valid measurement of parenting and family functioning within large longitudinal studies presents a number of challenges. As described earlier, parenting is a complex, multi-dimensional construct, the quality of which may be assessed by examining the emotional or affective tone of the interactions between parent and child, the frequency of specific behaviours (positive and negative), and the relative patterning of interactions and activities undertaken by parents with their children. Measurement protocols that use independent observers to code parent–child interactions under naturalistic or experimental conditions are considered the gold standard in studies of parenting and its effects on children (Webster-Stratton & Lindsay 1999). However, the time and burden costs of such approaches generally preclude their use in population studies. As an alternative, a number of parent self-report instruments developed and refined over time have proven to be valid proxies for direct observation methods (Lovejoy 1991; Lovejoy et al. 1999; Webster-Stratton & Lindsay 1999).
However, participant burden can still remain an issue for parent self-report measures, with multiple items being required to provide valid assessments for each dimension of parenting. This raises the dilemma for researchers of the need to balance data collection burden (by using a reduced set of items) with the goals of data collection validity (generally requiring longer, more comprehensive measures). Additionally, the validity and appropriateness of self-report measures may be culturally specific. Even within westernised societies, self-report instruments developed in one country or culture may not translate well to another country or culture. These measurement challenges apply equally to the measurement of parenting and to the measurement of indicators of family functioning (Nicholson et al. 2006b).
Finally, it is important to comment on the measurement challenges that arise when assessing general population samples as opposed to clinical samples of parents or families presenting for the treatment of an identified problem (for example, child conduct problems, child depression, marital distress, family conflict or violence). Many of the self-report tools available for assessing parenting and family functioning have been developed through research conducted with families experiencing significant problems. Such measures may discriminate well for those with high rates of problems, but show restricted or skewed distributions when applied to general population samples (Nicholson et al. 2006b).
Taking these considerations into account, a number of parent self-report measurement tools were selected for assessing parenting within the two LSAC cohorts, which:
- aimed to provide reliable and valid measurement at a population level
- comprised a relatively small number of items
- assessed key parenting dimensions that were developmentally relevant and broadly culturally appropriate for both age cohorts.
Similar criteria were applied to the selection of measures of family functioning, including the measurement of:
- parenting self-efficacy, parental mental health, stress and coping
- inter-parental conflict, support and relationship quality
- the supports provided by external family members.
Additionally, where possible, measures were selected that would enable comparisons between the LSAC cohorts and other relevant (preferably Australian) data sets (Nicholson et al. 2006b; Sanson et al. 2002).
Details of the specific measures employed (including actual items where appropriate) are provided in
subsequent sections.
1.8 Description of the LSAC sample
This section describes the LSAC sample used for the data analyses in this paper. The data were from data release 2.4 for Wave 1, which included 5,107 infants and 4,983 children.
Families with infants (infant cohort)
Infant characteristics
The average age of the infant cohort at the time of interview was 8.8 months (range: 3 to 19 months) and 51 per cent were male. The age of infants at interview extended beyond the upper limit for the target age range (that is, 12 months) due to the time lags between the selection of the sample and when contact was achieved and interviews conducted.
Carer characteristics
The primary carer (P1), the most knowledgeable person about the child, was almost always the child’s biological parent (99.8 per cent) and female (98.6 per cent). Table 1 summarises their characteristics. Just over 60 per cent were in the age range 25 to 34 years at the time of the birth of the child. Only 16.3 per cent of primary carers were 24 years of age or less. Over three‑quarters of primary carers were born in Australia. About 30 per cent had obtained some level of university education, 10.4 per cent had completed education to Year 10, with another 18.4 per cent completing Years 11 and 12. A small proportion of primary carers (4.8 per cent) had Year 9 or less education. Half of all primary carers were not in the labour force, another 13 per cent were employed full time and 34 per cent were employed part time, while 3.4 per cent reported being unemployed.
Over half (55.3 per cent) of secondary carers (P2) were in the age range 25 to 34 years at the time of the birth of the child and another 37.7 per cent were 35 years or older (Table 2). As with the primary carer, secondary carers were predominately born in Australia (76.4 per cent), just over one in four (27.7 per cent) had obtained some level of university education and 7.6 per cent and 13.6 per cent had achieved Year 10 education or completed Years 11 and 12 respectively. Most secondary carers (85.6 per cent) were in full time employment while 3.2 per cent reported that they were unemployed.
Family characteristics
Families of infants were distributed across Australia with sampling in proportion to population. About one-third of families lived in New South Wales, a one-quarter in Victoria and about one-fifth in Queensland (refer to Table 3). About 81 per cent of families with infants were intact couple families (that is, all children in the family being biological offspring of both parents), 10.5 per cent were lone-parent families, while 6.8 per cent were step-families (that is, at least one child in the family being a step-child of either parent). One-fifth of families earned less than $600 per week and about 13 per cent earned more than $2,000 per week. Three-quarters of the infants were either the only child in the family (39.1 per cent) or had one sibling (36.4 per cent).
| n=5,107 | Estimated % | Mean | Standard error |
|---|---|---|---|
| Gender | |||
| Male | 1.4 | ||
| Female | 98.6 | ||
| Age at child’s birth (years) | 30.3 | 0.130 | |
| Less than 25 | 16.3 | ||
| 25–34 | 61.5 | ||
| 35 and above | 22.2 | ||
| Primary carer place of birth | |||
| Australia | 78.6 | ||
| Asia | 7.4 | ||
| United Kingdom | 4.1 | ||
| New Zealand | 3.3 | ||
| Africa or Middle East | 3.1 | ||
| Other Europe | 1.6 | ||
| Other Oceania | 1.2 | ||
| North or South America | 0.7 | ||
| Education | |||
| Less than Year 9 | 4.8 | ||
| Year 10 | 10.4 | ||
| Years 11–12 | 18.4 | ||
| Trade certificate or diploma | 37.2 | ||
| University | 29.2 | ||
| Employment | |||
| Full-time employment | 13.0 | ||
| Part-time employment | 34.4 | ||
| Unemployed | 3.4 | ||
| Not in labour force | 49.1 |
| n=4,571 | Estimated % | Mean | Standard error |
|---|---|---|---|
| Gender | |||
| Male | 98.3 | ||
| Female | 1.7 | ||
| Age at child’s birth (years) | 33.1 | 0.121 | |
| Less than 25 | 7.0 | ||
| 25–34 | 55.3 | ||
| 35 and above | 37.7 | ||
| Primary carer place of birth | |||
| Australia | 76.4 | ||
| Asia | 7.1 | ||
| United Kingdom | 6.0 | ||
| New Zealand | 3.4 | ||
| Africa or Middle East | 3.0 | ||
| Other Europe | 2.1 | ||
| Other Oceania | 1.2 | ||
| North or South America | 0.7 | ||
| Education | |||
| Year 9 or less | 3.4 | ||
| Year 10 | 7.6 | ||
| Years 11–12 | 13.6 | ||
| Trade certificate or diploma | 47.7 | ||
| University | 27.7 | ||
| Employment | |||
| Full-time employment | 85.6 | ||
| Part-time employment | 7.0 | ||
| Unemployed | 3.2 | ||
| Not in labour force | 4.3 |
| n=5,107 | Estimated % |
|---|---|
| State of residence | |
| New South Wales | 33.7 |
| Victoria | 25.4 |
| Queensland | 19.1 |
| Western Australia | 9.6 |
| South Australia | 7.0 |
| Tasmania | 2.4 |
| Australian Capital Territory | 1.7 |
| Northern Territory | 1.1 |
| Family type | |
| Lone parent | 10.5 |
| Married, intact | 66.5 |
| De facto, intact | 15.0 |
| Married, step | 4.0 |
| De facto, step | 2.8 |
| Other | 1.2 |
| Number of children(a) | |
| 1 | 39.1 |
| 2 | 36.4 |
| 3 | 16.4 |
| 4 or more | 8.1 |
| Weekly household income (combined P1 and P2)(a) | |
| Less than $600 | 20.2 |
| $600–$999 | 27.6 |
| $1,000–$1,499 | 25.9 |
| $1,500–$1,999 | 13.4 |
| More than $2,000 | 12.8 |
(a) Refer to Appendix A for more details.
Families with children (child cohort)
Child characteristics
The average age of the child cohort was 57 months (range: 51 to 67 months) and 51 per cent were male. Again, the age of children at interview extended beyond the upper limit for the target age range (that is, 60 months) due to time lags between the selection of the sample and when contact was achieved and interviews conducted.
Carer characteristics
As with the infant cohort, in the child cohort the primary carer was the most knowledgeable person about the child, of which almost all were the child’s biological parent (99.4 per cent) and female (97.3 per cent). Table 4 summarises their characteristics. Almost 65 per cent (64.5 per cent) were in the age range 25 to 34 years at the time of the birth of the child and 15.9 per cent of primary carers were 24 years of age or less. Over three quarters of primary carers were born in Australia. About 24 per cent had obtained some level of university education, 12.4 per cent had completed education to Year 10, with another 20.1 per cent completing Years 11 to 12. A small proportion of primary carers (6.1 per cent) had Year 9 or less education. Of all primary carers, 41 per cent were not in the labour force, another 15.4 per cent were employed full time, and 39.8 per cent were employed part- time, while 4.3 per cent reported being unemployed. Where primary carers reported being employed, this was for an average of 24.3 hours per week.
Over half (58.1 per cent) of secondary carers were in the age range 25 to 34 years at the time of the birth of the child and another 34.6 per cent were 35 years or older (Table 5). As with primary carers, secondary carers were predominately born in Australia (73.8 per cent), just over one in four (27.3 per cent) had obtained some level of university education and 8.4 per cent and 13.3 per cent respectively had achieved Year 10 education or completed Years 11 to 12. Most secondary carers (86.1 per cent) were in full-time employment while 2.1 per cent reported that they were unemployed.
Family characteristics
Families of children were distributed across Australia with sampling in proportion to population. About one third of families lived in New South Wales, a one-quarter in Victoria and about one-fifth in Queensland (Table 6). Intact couple families made up 75 per cent of families with children, 15 per cent were lone‑parent families, while 8.3 per cent were step-families. One-fifth of families earned less than $600 per week and about 15 per cent earned more than $2,000 per week.
Compared to the infant cohort, a smaller proportion of the child cohort were only children (11.5 per cent compared to 39.1 per cent). Nearly half of the households contained two children (47.5 per cent), 26.8 per cent had three children and 14.2 per cent had four or more children.
| n=4,983 | Estimated % | Mean | Standard error |
|---|---|---|---|
| Gender | |||
| Male | 2.7 | ||
| Female | 97.3 | ||
| Age at child’s birth (years) | 29.9 | 0.117 | |
| Less than 25 | 15.9 | ||
| 25–34 | 64.5 | ||
| 35 and above | 19.6 | ||
| Primary carer place of birth | |||
| Australia | 76.4 | ||
| Asia | 8.2 | ||
| United Kingdom | 5.6 | ||
| New Zealand | 2.6 | ||
| Africa or Middle East | 3.2 | ||
| Other Europe | 2.1 | ||
| Other Oceania | 1.3 | ||
| North or South America | 0.5 | ||
| Education | |||
| Year 9 or less | 6.1 | ||
| Year 10 | 12.4 | ||
| Years 11–12 | 20.1 | ||
| Trade certificate or diploma | 37.2 | ||
| University | 24.1 | ||
| Employment | |||
| Full-time employment | 15.4 | ||
| Part-time employment | 39.8 | ||
| Unemployed | 4.3 | ||
| Not in labour force | 40.5 |
| n=4,238 | Estimated % | Mean | Standard error |
|---|---|---|---|
| Gender | |||
| Male | 97.4 | ||
| Female | 2.6 | ||
| Age at child’s birth (years) | 36.6 | 0.130 | |
| Less than 25 | 7.3 | ||
| 25–34 | 58.1 | ||
| 35 and above | 34.6 | ||
| Primary carer place of birth | |||
| Australia | 73.8 | ||
| Asia | 7.4 | ||
| United Kingdom | 6.6 | ||
| New Zealand | 3.0 | ||
| Africa or Middle East | 4.0 | ||
| Other Europe | 3.2 | ||
| Other Oceania | 1.5 | ||
| North or South America | 0.6 | ||
| Education | |||
| Year 9 or less | 4.0 | ||
| Year 10 | 8.4 | ||
| Years 11–12 | 13.3 | ||
| Trade certificate or diploma | 47.0 | ||
| University | 27.3 | ||
| Employment | |||
| Full-time employment | 86.1 | ||
| Part-time employment | 6.6 | ||
| Unemployed | 2.1 | ||
| Not in labour force | 5.1 |
| n=4,983 | Estimated % |
|---|---|
| State of residence | |
| New South Wales | 34.2 |
| Victoria | 24.5 |
| Queensland | 19.6 |
| Western Australia | 9.6 |
| South Australia | 7.1 |
| Tasmania | 2.5 |
| Australian Capital Territory | 1.6 |
| Northern Territory | 0.9 |
| Family type | |
| Lone parent | 15.0 |
| Married, intact | 67.6 |
| De facto, intact | 7.4 |
| Married, step | 5.1 |
| De facto, step | 3.2 |
| Other | 1.7 |
| Number of children(a) | |
| 1 | 11.5 |
| 2 | 47.5 |
| 3 | 26.8 |
| 4 or more | 14.2 |
| Weekly household income (combined P1 and P2)(a) | |
| Less than $600 | 19.3 |
| $600–$999 | 24.8 |
| $1,000–$1,499 | 24.7 |
| $1,500–$1,999 | 16.5 |
| More than $2,000 | 14.7 |
(a) Refer to Appendix A for more details.
1.9 Discussion
The study children and their families were similar across both cohorts in several respects. Between the two cohorts the proportions of study children who were male or female were similar (51 per cent male), as were the proportions of primary and secondary carers who were female and male respectively (approximately 98 per cent), and Australian born (approximately 76 per cent). Similar distributions were obtained for each cohort in terms of primary and secondary carer age at the time of the child’s birth, and highest levels of educational attainment. The households were also similar across the two cohorts in geographical location (measured at the level of state or territory) and distributions for total household income.
As expected for cohorts of different ages, there were also a number of demographic differences between the cohorts. Compared to the infant cohort the total number of children in the families of the child cohort was greater. For the infant cohort 39.1 per cent of infants were the only child in the household, while this was the case for only 11.5 per cent of the child cohort. Workforce participation was also different between the cohorts for the primary carer. While 47.4 per cent of primary carers for the infant cohort were currently in employment, 55.2 per cent of primary carers for the child cohort were currently employed, although the proportion of those employed who were full time was similar for both (around 27.5 per cent). Workforce participation for secondary carers was similar for both cohorts, with around 86 per cent in current full-time employment. It is likely that these demographic differences are due to the age of the child, and hence the stage of family life cycle at the time of interview. The extent to which this is the case will become more apparent once subsequent waves of longitudinal data have been collected.
Finally, as was noted in the methods section, response rates for the cohorts were 64 per cent for the infants and 57 per cent for the 4 to 5 year olds. This introduces the potential for sampling biases. Indeed cohort members with primary carers who had completed Year 12 were overrepresented in the final sample, infants with no siblings were underrepresented, and children in couple families were overrepresented, while those in lone-parent families were underrepresented. These differences were addressed through sample weighting (Soloff et al. 2006). However, on a wide range of other demographic data the LSAC cohorts were found to be similar to the age-matched Australian population.