GLOSSARY

Apgar Score

A score, determined by the five characteristics of colour, muscle tone, heart rate, respiration and reflex activity, which is assigned to newborn babies at one and five minutes after birth.

Appagrent Retention Rate

The apparent retention rate is defined as the percentage of full-time students of a given cohort group (i.e. all students who commence secondary school in the same year) who continue from the first year of secondary schooling to a specified year at school. The apparent retention rate can exceed 100 per cent, as it does not account for the transfer of students between jurisdictions, migration, or students repeating a year.

Australian Qualifications Framework

The Australian Qualifications Framework (AQF) is a unified framework for classifying all qualifications in post-compulsory education and training in Australia.1 It covers schools, vocational education and training (TAFEs and private providers) and the higher education sector (mainly universities). The second edition of the Australian Standard Classification of Occupations (ASCO) assigned each major occupation group to one of five broad skill levels. These skill levels were defined in terms of the levels of qualifications as defined in the AQF that are generally required for the occupations in each group and years of relevant work experience.

The ASCO major occupation groups have been grouped into skill levels as described in the following table.²

(a) In some instances relevant experience is required in addition to the formal qualification.

Carer Education

The level of educational attainment achieved by carers was determined from two survey questions: the highest grade finished at school, and what post-school qualifications have been achieved. Post-school qualifications were classified as:

Trade/apprenticeship

Certificate from college

Diploma (beyond year 12)

Bachelor Degree

Post Graduate Diploma/higher degree

Other

Carers who had completed a diploma, bachelor degree, post graduate diploma or higher degree were classified as having 13+ years education. Otherwise educational attainment was classified by highest grade finished at school. The following categories have been used in this publication:

Did not attend school

1-9 years education

10 years education

11-12 years education

13+ years education

Note that educational attainment refers to highest level achieved, not the number of years taken to achieve the qualification.

Community Housing

Carers who reported that their dwelling was rented were asked if their dwelling was a private rental, or rented from HomesWest, the Aboriginal Housing Authority or Community Housing. HomesWest and the Aboriginal Housing Authority are two schemes administered by the Department of Housing and Works, with HomesWest being the general public housing scheme for WA.

Community Housing generally refers to housing that is provided in discrete Aboriginal Communities where the housing comes under the control of an Aboriginal Housing Organisation. Funding for community housing organisations comes from a combination of Commonwealth and State government sources. One key difference with Community Housing is the degree of direct Aboriginal involvement in the planning and delivery of housing through Aboriginal Housing Organisations.

Dietary quality indicators

Carers were asked a number of questions relating to the diet of the Aboriginal children in their care, including information about how often children ate fruit and vegetables, and what types of beverages were consumed. The available data allowed a range of indicators of dietary quality to be devised. These indicators did not measure dietary intake, but were designed to reflect whether the principles of a healthy diet were being observed. It must be noted that these indicators are based on interview responses, which were not further validated.

Indicator 1: met if water was usually drunk when thirsty.

Indicator 2: met if some form of unsweetened and unflavoured cow or soy milk was regularly consumed.

Indicator 3: met if fresh fruit was usually consumed on 6 or 7 days of the week.

Indicator 4: met if at least half a cup of a variety of at least three fresh vegetables, other than potato, were usually consumed on 6 or 7 days of the week.

The number of these indicators that were met was used as an overall indicator of dietary quality.

Dwellings

In household surveys a distinction is often made between dwellings, households and families as per the Census of Population and Housing, with allowance made for the possibility of more than one household living in a single dwelling, and for a household to comprise more than one family. In the census, a dwelling is a habitable structure, a household is a group of related or unrelated people who make common provision for food, while a family is a group of people related by blood, marriage, adoption, step or fostering who usually reside within a single family. Note that in a block of flats, for example, each flat is considered to be a separate dwelling.³

In practice, the distinction between dwellings, households and families was found to have little importance in the WAACHS. Aboriginal families living together often contain extended family relationships. However, there were hardly any cases where two or more unrelated families were found to be living in the same household and no cases were found where multiple households were residing in the same dwelling. In this volume, the terms household and family are used interchangeably, while the term dwelling is used to describe the physical structure in which a household or family is living.

Note that the survey only included private dwellings. Non-private dwellings, such as hotels or boarding schools, were not included in the scope of the survey. However, carers were asked if there were any children aged under 18 years who usually live at this address but who are temporarily away. Information about these children was collected from the carers where appropriate.

Family Functioning

Family disharmony is known to be associated with poorer child development outcomes. The survey used a 9-item scale to measure the extent to which families have established a climate of trust and cooperation, emotional support and good communication. Primary carers were asked to rate each of nine statements on a scale of 1–5 as to how accurately each statement described their family circumstances. The nine statements included items about communications and decision making in the family, emotional support, time spent together, and family cooperation. These ratings were summed to produce an overall score. Families were then split into quartiles based on this score, with approximately 25 per cent of children in each category. These categories have been labelled poor, fair, good and very good family functioning in this publication. For details of the nine items and how they were combined to form the family functioning score, see Appendix C of Volume Two⁴ — Measures derived from multiple responses and scales.

Family Trees

One of the first tasks for interviewers at the beginning of each interview was the completion of the Household Record Form which included listing all usual residents of the household and the relationships between them. To assist in describing these relationships interviewers also drew a family tree to summarise relationships within the household. The Household Record Forms were used in conjunction with the family trees to classify the household structure and the family care arrangements for each child.

Household carer

The term ‘household carer’ is used when referring to analyses at the dwelling level, as opposed to the more common primary carer-level or child-level analyses presented in this volume.

For analyses at the dwelling level a single response per dwelling was required. There were some instances where more than one primary carer was living in the same dwelling. In order to exclude multiple assessments of individual dwellings in these instances, the analyses in Chapter Six are restricted to assessments by one carer per dwelling.

A single assessment for each household was achieved by nominating a ‘household carer’, whose assessment of the housing items was used to analyse dwelling level outcomes. Therefore analysis at the dwelling level refers to 11,400 ‘household carers’ as opposed to the usual 12,600 ‘primary carers’ used elsewhere in this volume. See comment box entitled Analysis of household and dwelling level data in this chapter in Chapter Six of Volume 4 Strengthening the Capacity of Aboriginal Children, Families and Communities⁵ for a more complete explanation.

Household Care Arrangements

Two classifications have been used to describe the structure of each household. The first described the overall household structure (See household composition below). The second classification describes the care arrangements for each child and has been assigned at the child level. In many cases both classifications will be equivalent. For instance, for a family with two original parents and two children the household composition would be classified as “Two parent household - nuclear type”, and the care arrangement for each child would be classified as “Both original parents”. However, for a blended household with, say, two parents and one child, plus a child from a previous union, the care arrangements for each child would not be the same. For one child the care arrangement would be classified as “Both original parents”, while for the other child the care arrangement would be classified as “one parent and new partner”.

In the case of extended families, where say Aunts and Uncles, Grandparents and other relatives are living in the household, the household composition classification would describe all the usual residents of the household. In terms of the care arrangements for each child, extended family relationships are only classified as part of the care arrangements for the child if the extended family member was involved in caring for the child.

Household Composition

Two separate classifications have been developed for describing the structure of each household. The first describes the composition of the household. Households were classified based on the information recorded on the Household Record Forms and the family trees that were drawn by interviewers describing each family. The household classification discriminates between nuclear and extended families and describes the generational complexity of extended families.

Household Occupancy Level

A two-level index of household occupancy was created based on the number of bedrooms and the number of people usually sleeping in the home. A household was considered to have a high level of household occupancy if it had the following attributes in terms of the number of bedrooms and the number of people sleeping in the home.

Number of bedrooms		Number of people sleeping there
1		5 or more
2		6 or more
3		7 or more
4		8 or more
5 or more		9 or more

Note that the definition of household occupancy level published on page 129 of Volume Two was incorrect. The above definition has been used consistently throughout all analysis of the survey data.

Household Record Form

In addition carers were asked if there were any children under 18 years who usually live at this address but who were temporarily away. For example, if a child was away at boarding school, information about that child would still be collected from the carer.

ICC Regions

With the abolition of ATSIC Regional Councils and the establishment by the Office of Indigenous Policy Coordination (OIPC) of regional Indigenous Coordination Centres (ICCs), changes may be made to the geographic regions used for producing statistics in relation to Aboriginal peoples. In Western Australia, seven ICCs have been established. The boundaries of the regions served by these ICCs are similar to those of the nine ATSIC regions in Western Australia, with the Perth Noongar and Noongar Country (Narrogin) ATSIC regions combined into the Perth ICC region, and the Western Dersert (Warburton) and Mulga Malliee (Kalgoorlie) ATSIC regions combined into the Kalgoorlie ICC region.

For the purposes of these publications, it is assumed that the boundaries of the nine former ATSIC regions will remain unchanged, and these regions are now referred to as ICC regions.

Index of Relative Socio-Economic Disadvantage

The index of relative socio-economic disadvantage is one of five measures of socio-economic status calculated by the ABS in their SEIFA product.⁶ The index is a summary measure calculated from census data which ranks the relative level of disadvantage of each census collection district (CD). As one of the factors included in the standard SEIFA product is proportion of Aboriginal and Torres Strait Islander people in each CD, the ABS produced a special version of the index for use in this survey that excluded this variable as a factor. The index is scaled to have a mean of 1000 and a standard deviation of 100. Lower values indicate greater levels of disadvantage

Indigenous Status

To be included in the survey, carers had to identify their children as being of Aboriginal or Torres Strait Islander origin. Only Aboriginal or Torres Strait Islander children (under the age of 18 years) were included in the survey, even in those cases where there were both Aboriginal and non-Aboriginal children living in the same household. Note that the carers did not have to be Aboriginal for the family to be included in the survey.

Carers were also asked whether they were of Aboriginal or Torres Strait Islander descent. Approximately 17 percent of primary carers and 21 percent of secondary carers of Aboriginal and Torres Strait Islander children and young people were not of Aboriginal or Torres Strait Islander descent.

Level of Relative Isolation (LORI)

A new classification of remoteness and isolation has been introduced in this survey – the Level of Relative Isolation (LORI). The LORI is based on a recently introduced product from the National Key Centre for Social Application of Geographic Information Systems at Adelaide University (GISCA) called ARIA++. The ARIA++ is an extension of ARIA (the Accessibility/Remoteness Index of Australia), which was first published in 1997 and has been widely adopted as the standard classification of remoteness in Australia. Because ARIA is based on describing the entire population of Australia, it has not been specifically designed to describe the circumstances of Aboriginal people living in remote areas. The ARIA++ gives much greater discrimination among more remote areas by including more service centres, of smaller sizes, in calculating its remoteness scores.

Based on the ARIA++ scores, five categories of isolation have been defined specifically for the survey that reflect differences in access to services for Aboriginal children. To avoid confusion with the original ARIA, the five categories are referred to as Levels Of Relative Isolation (LORI) and range from None (the Perth Metropolitan area) to Low (eg Albany), Moderate (eg Broome), High (eg Kalumburu) and Extreme (eg Yiyili).

See A New Way of Looking at Remoteness and Isolation in Chapter 1, and Appendix C - Determination of Levels of Relative Isolation from ARIA++ for more details.

Life stress events

The number of life stress events that occur in a single period can impact on a families’ abilities to cope. Most people are able to cope with a single stressful event, but when multiple stressful or traumatic events occur simultaneously or over a relatively short time period it can be more and more difficult to cope.

In the WAACHS, primary carers were asked if any of fourteen major life stresses had occurred in the family in the preceding twelve months. These events included illness, hospitalisation or death of a close family member, family break up, arrests, job loss and financial difficulties.

For analysis, the number of life stress events in the previous 12 months were grouped as follows: 0–2, 3–4, 5–6, 7–14, with each category containing approximately one-quarter of survey children.

For details of the life stress events measured in the survey see Appendix C of Volume Two — Measures derived from multiple response scales. This volume can be downloaded free from our website: www.ichr.uwa.edu.au/waachs.

It has been possible to compare the experience of life stress events in families with non-Aboriginal children and in families with Aboriginal children as thirteen of the life stress events asked in the WAACHS were the same items (or with minor wording variations) as was asked in the 1993 Western Australian Child Health Survey.⁷

Logistic regression

See Multivariate Logistic Regression Modelling

Main Language spoken

Teachers of surveyed school students were asked the main language spoken by the child in the classroom, playground and at home. Students’ main language spoken was classified into five categories:

English

Aboriginal English

Kriol/Creole

Aboriginal language

Other (specify).

Malcolm defines Aboriginal English as ‘a range of varieties of English spoken by many Aboriginal and Torres Strait Islander people and some others in close contact with them which differ in systematic ways from Standard Australian English at all levels of linguistic structure and which are used for distinctive speech acts, speech events and genres.’ ⁸

In communities which brought together Aboriginal people from a number of mutually unintelligible languages, complex new languages known as creoles developed to allow children who grew up speaking pidgin as a first language to communicate. In northern areas of Australia, many Aboriginal people speak a creole language such as Kriol as well as other languages.⁸

Multivariate Logistic Regression Modelling

Logistic regression is a modelling technique that is used to investigate the relationship between the probability of a certain outcome (for example, a child having a particular health condition) and a set of explanatory variables. Logistic regression is discussed in several statistical publications – see, for example, Hosmer and Lemeshow (2000).⁹ In this publication, logistic regression models have been fitted using a weighted version of multi-level modelling which allows for community level, family level and individual level factors to be included as explanatory variables in the models (see Pfeffermann et al, 1997).¹⁰

This technique takes into account the survey weights and the hierarchical structure of the data with selection of children within families and communities.

Logistic regression modelling has been used in situations where multiple factors may all have an impact on an outcome of interest. If the factors themselves are inter-related, cross-tabulation analysis may not tell the full story. For each variable included in a logistic regression model, the model determines its effect on the outcome independent of the effect of all other variables included in the model.

Odds Ratio

The odds of a given event is the ratio of the probability of its occurrence to the probability of its non-occurrence. For instance the odds of obtaining heads in a coin toss are one to one, the odds of any given face in the roll of a die are one to five. The odds ratios used in this publication are a measure of relative risk, derived from a formula which examines the association between, in most of the survey cases, a risk factor (exposure), and an adverse health outcome. In this publication odds ratios have been estimated using logistic regression which estimates the effect of each risk factor included in a model after adjusting for the independent effects of all other factors included in the model. The statistical significance of an odds ratio can be judged by whether the confidence interval includes the reference value of one.

Outstations

Generally speaking outstations are small Aboriginal communities where families live in close connection with the natural environment. These outstation communities are often linked to a larger parent Aboriginal community for the provision and maintenance of services.

Percentage of optimal birthweight (POBW)

An infant’s weight at birth depends on both the length of gestation and the rate at which it has grown in utero. Not all foetuses grow at the same rate. Boys grow faster than girls, children of tall mothers grow faster than those of short mothers, and a women’s first child grows more slowly than her subsequent children. However growth rate is also affected by a number of pathological conditions, most of which decrease growth rate (the exception being maternal diabetes, which increases growth rate). The appropriateness of an infant’s growth can be estimated as the ratio of the infant’s observed birth weight to the infant’s optimal birthweight. Infants that have grown normally have a POBW close to 100 per cent and, in these analyses, percentages below 85 per cent are classified as having sub-optimal intrauterine growth.¹¹

Primary Carer

For each child in the survey, the family was asked to identify the primary carer of that child. This was the person who was considered to spend the most time with the child or who had primary responsibility for the upbringing of the child. In many cases, the primary carer was the child’s mother. The primary carer was then asked to provide information about each of their children for the survey.

Quality of Parenting

The nature of the relationship between a child and his or her primary carer, and the style and quality of the carer’s parenting are important influences on the development and wellbeing of children. The survey asked a series of questions of carers about their relationship with each of their children. An index of quality of parenting has been derived from three of these items: how often carers praise their children, how often they hit or smack their children and how often they laugh together with their children. These three items, which measure the concepts of parenting warmth and harshness, were rated by carers on a five-point frequency scale from ‘Never’ through to ‘Almost always’. An overall score was produced by summing these three items. Children were then ranked by score, and split into quartiles based on this score, with approximately 25 per cent of children in each category. These categories have been labelled poor, fair, good and very good quality of parenting in this publication.

For further details on the quality of parenting items, and how they were combined to form the quality of parenting score, see Appendix C — Measures derived from multiple responses and scales of Volume Two, The Social and Emotional Wellbeing of Aboriginal Children and Young People⁴.

Record Linkage

Carers were asked for consent to access their hospital and medical records, as well as the birth, hospital and medical records of their children. Carers who consented were given the opportunity to opt out at any stage should they change their mind. The vast majority of carers consented to these records being accessed. Of primary carers, 96.7 per cent consented to allow access to their hospital records, while 92.8 per cent of secondary carers gave similar consent. Overall, 96.3 per cent of carers gave consent for their children’s birth, hospital and medical records to be accessed.

The WA Record Linkage System is unique in Australia, and one of only a handful of similar data collections in the world. It links together birth and death registrations with administrative hospital data from several sources to give a comprehensive record of health services contacts for the population of Western Australia. As there are no unique identifying numbers, probabilistic record linkage has been used to link the files together. This operates on matching names, dates of birth, hospital names and addresses. The procedure allows for possible changes in the matching fields by calculating the probabilities of records being correct matches. Records that are potential links are clerically reviewed, and the overall error rate has been estimated to be less than one per cent.

Key components of the record linkage system used in the survey are the birth records, the Hospital Morbidity Data System and the Mental Health Information System.

Relative Risk

Relative risk is a measure of how much a particular factor influences the risk of a particular outcome. It is calculated as the ratio of the proportion of a particular group of individuals that have a condition to the proportion with the same condition in a reference group of individuals. In this survey the reference group is usually either the total WA population in the same age range, or the non-Aboriginal population in the same age range.

Secondary Carer

Each family was asked to identify the primary and secondary carer of each child. Often the secondary carer was the father of the child, but may also have been a grandparent or other relative of the child, or other person involved in the upbringing of the child.

SocioEconomic Index for Schools

The Socioeconomic Index for schools (SEI) is an index of socioeconomic disadvantage which is constructed mainly from data collected by the ABS at the latest census. The SEI has five dimensions: Education, Occupation, Aboriginality, Single Parent Family, and Family Income, with the first three being double weighted, the last two single weighted. The Education, Occupation, and Single Parent Family dimensions are based on ABS Census data, the Aboriginality dimension is based on the proportion of Aboriginal students in the school, and the Family Income dimension is based on ABS Census data on the income of families with school-aged children adjusted according to the Regional Price Index. The effect of the Regional Price Index adjustment is to reduce the dimension values in districts where prices are higher, especially in the Kimberley and Pilbara.

The dimension scores are constructed for each census collection district (CD) first, by a series of principal component analyses of the CD level census data. These are then standardised with a mean of 100 and standard deviation of 10, and then school scores are constructed from them. The proportion of Aboriginal students in each school is converted into a similar form.

The addresses of students are collected from schools. These addresses are then mapped to the collection districts, and the proportion of students in each CD is calculated for each school. This enables the school SEI to be calculated from the standardised CD dimension scores, as follows:

Education (x 2)

Occupation, including unemployment (x 2)

Aboriginality (x 2)

Single Parent Family

Family Income, adjusted by the Regional Price Index for the district.

The SEI is a variation of the type of index developed by Kenneth Ross and Stephen Farish, and the analysis is still performed by Professor Stephen Farish who is now at the University of Melbourne.

Strengths and difficulties questionnaire

In this survey, the Strengths and Difficulties Questionnaire (SDQ) was used to measure emotional or behavioural difficulties in Aboriginal children. The SDQ comprises twenty-five questions looking into five areas of emotional and behavioural difficulties: emotional symptoms, conduct problems, hyperactivity, peer problems and prosocial behaviour. The responses from the twenty questions related to the first four of these areas are combined to produce the Strengths and Difficulties Total Score. This score can range from zero to a maximum score of 40.

Information about the emotional and behavioural difficulties of Aboriginal children was collected from three sources: their primary carer, school teacher, and young people aged 12–17 years themselves. In this publication, most of the analysis of Aboriginal children’s emotional and behavioural difficulties are based on teacher reported SDQ.

The Strengths and Difficulties Total Score can be grouped into three ranges — the normal range (0–11), borderline range (12–15) and abnormal range (16–40). These categories and their ranges are described by Goodman.12

Classification of the SDQ Total Score into normal, borderline and abnormal ranges is typically used within a clinical setting by mental health professionals to help identify and diagnose specific emotional or behavioural difficulties among children. In clinical settings, the SDQ may be used in conjunction with other techniques to assess an individual child in accordance with recognised diagnostic standards.

In household-based population surveys such as the WAACHS, where it is not possible to conduct comprehensive clinical assessments of individual children, the SDQ is more appropriately used to assess risk status for clinically significant emotional or behavioural difficulties. Thus, groups of children with SDQ scores in the range:

0–11 are identified as having low risk of clinically significant emotional or behavioural difficulties

12–15 are identified as having moderate risk

16–40 are identified as having high risk.

As described in Goodman, 12 the cut-offs used to assess risk of clinically significant emotional or behavioural difficulties are slightly different when carers of the child complete the SDQ. Volume Two of the WAACHS contains an extensive analysis of carer reported emotional and behavioural difficulties.

Tenure Type

Tenure type describes the legal right of a household to occupy a dwelling (eg fully owned, being purchased, rented or some other arrangement).

Usual Residents

The survey was conducted with interviewers going from door to door in search of Aboriginal families. Each family was asked “Are there any Aboriginal children or teenagers living at this address who are aged between 0 and 18 years?” Families answering yes to this question were included in the scope of the survey. Thus the survey was based on carers identifying their children as being of Aboriginal or Torres Strait Islander origin. Note that the carers need not be Aboriginal for the family to be included in the survey.

The names of all people who usually live at the dwelling were listed on the Household Record Form. In addition carers were asked if there were any children under 18 years who usually live at this address but who were temporarily away. For example, if a child was away at boarding school, information about that child would still be collected from the carer.

Endnotes

1 Australian Qualifications Framework Advisory Board (2002). Australian Qualifications Framework: Implementation Framework. 3rd Edition. Melbourne: AQF Advisory Board.

2 Australian Bureau of Statistics (1997). Australian Standard Classification of Occupations Second Edition. ABS Cat. No. 1220.0. Canberra: Australian Bureau of Statistics.

3 Australian Bureau of Statistics (2001). 2001 Census Dictionary. ABS Cat. No. 2901.0. Canberra: Australian Bureau of Statistics.

4 Zubrick SR, Silburn SR, Lawrence DM, Mitrou FG, Dalby RB, Blair EM, Griffin J, Milroy H, De Maio JA, Cox A, Li J. The Western Australian Aboriginal child health survey: The social and emotional wellbeing of Aboriginal children and young people. Perth: Curtin University of Technology and telethon Institute for Child Health Research; 2005.

5 Silburn SR, Zubrick SR, De Maio JA, Shepherd C, Griffin JA, Mitrou FG, Dalby RB, Hayward C, Pearson G. The Western Australian Aboriginal Child Health Survey: Strengthening the Capacity of Aboriginal Children, Families and Communities. Perth: Curtin University of Technology and telethon Institute for Child Health Research; 2006.

6 Australian Bureau of Statisics. Information Paper. 1996 Census of Population and Housing. Socio-economic Indexes for Areas. ABS Cat. 2039.0. Canberra: Australian Bureau of Statistics

7 Silburn SR, Zubrick SR, Garton AF, Burton P, Dalby R, Carlton J, Shepherd C, Lawrence D. Western Australian Child Health Survey: Family and community health. Perth: Australian Bureau of Statistics and the TVW Telethon Institute for Child Health Research; 1996

8 Malcolm IG. Language and communication enhancement for two-way education. Perth: Edith Cowan University; 1995.

9 Hosmer D, Lemeshow S. Applied logistic regression 2nd edition. New York: Wiley; 2000.

10 Pfeffermann D, Skinner CJ, Holmes DJ, Goldstein H, Rasbash J. Weighting for unequal selection probabilities in multi-level models. Journal of the Royal Statistical Society, Series B 1998;60:23–40.

11 Blair E. Why do Aboriginal neonates weigh less? II. Determinants of birthweight for gestation. Journal of Paediatrics and Child Health 1996;32.

12 Goodman R, Ford T, Simmons H, Gatward R, Meltzer H. Using the Strengths and Difficulties Questionnaire (SDQ) to screen for child psychiatric disorders in a community sample. British Journal of Psychiatry 2000;177:534–9
Goodman R. SDQ: Scoring the SDQ. [Online] [cited 2005 Oct 28]; Available from: URL: http://www.sdqinfo.com/ba3.html