All estimates have been derived from the survey sample, and weighted to represent the population of children and families in Australia. The estimates have been rounded, generally to three significant digits. As such, discrepancies may occur between the sums of component items and their totals.
Estimates have been suppressed where there were less than 5 survey respondents contributing to a cell. These appear as blank cells in tables and are omitted from charts.
Estimates are shown along with 95% confidence limits. Because Young Minds Matter was based on a sample and not a full census of Australian children and adolescents, the survey estimates could differ from the results that would be obtained from a full census due to random chance. The 95% confidence limits give an indication of the degree of sampling variability associated with the estimates. As a general rule, estimates that are based on larger numbers of contributing children are more accurate than those where fewer sample children contribute to the estimate. For instance, using the remoteness classification, the prevalence estimates for Major Cities are more accurate than those for Remote Australia because a higher proportion of the sample was located in Major Cities.
As an approximate rule of thumb, when comparing two prevalence estimates, if the ranges for the confidence limits of both estimates overlap, there is a higher chance that any difference between the two figures could be attributed to chance variation. When the ranges of the confidence limits do not overlap there is a greater chance that a full census would also have found a difference in the figures being compared.
Exhibiting symptoms of the problem in the 12 months prior to interview. Prevalence includes both new cases, whose symptoms first developed during the 12 months prior to the interview and continuing cases whose symptoms were present prior to the 12 months, but persisted, and were at a level to meet the diagnostic criteria in the 12 months prior to interview.
The survey did not assess DSM-IV criteria for eating disorders. Instead, the survey sought to identify eating behaviours that may be on the pathway to eating disorders. These were eating behaviours associated with low weight, and binge eating and purging.
Questions were taken from the Avon Longitudinal Study of Parents and Children to assess a range of activities young people may undertake to control their weight in the past 12 months. These were:
Young people also self-reported their height and weight, and this was used to assess their Body Mass Index (BMI). BMI ranges specifically designed for children and young people were used to classify underweight and overweight status.
Young people were considered to have low-weight problem eating behaviour if their BMI was in the underweight range and they dieted, fasted, vomited or used laxatives to lose weight or regularly exercised when they were supposed to be doing other things.
Young people were considered to have binge eating and purging problem eating behaviour if their BMI was not in the underweight range and they reported both binge eating and either vomiting or taking laxatives to control weight.
Approximately 8% of young people did not provide either their height or weight. These young people were excluded from the calculations of BMI and low weight problem eating behaviour and binge eating and purging problem eating behaviour.
Families were classified into families with two parents or carers and families with one parent or carer. Families with two parents or carers were further categorised into original, step, blended or other families corresponding to the Australian Bureau of Statistics family blending classification variable introduced in the 2006 Census. These are defined as follows:
Household income has been split into three approximately equally sized groups. Around 4% of families either didn’t know or refused to provide their household income. These families have been excluded from tables and charts relating to household income.
This classifies people as employed when working full-time, part-time or away from work, or not in employment when unemployed or not in the labour force. Employed includes casual, temporary or part-time work if it was for an hour or more.
For the purposes of the survey this was collected for both parents and carers for the previous week.
The index of relative socio-economic disadvantage is produced by the Australian Bureau of Statistics from the 2011 Census of Population and Housing, and gives a summary measure of the relative socio-economic disadvantage of the Statistical Area 1 (SA1) that the household is located in.
Area of residence was categorised as either Greater Capital Cities or Rest of state based on the Australian Bureau of Statistics Greater Capital City Statistical Area (GCCSA) classification. This classification represents the functional extent of the eight state and territory capital cities in Australia. Households within these areas were classified as Greater Capital Cities. The remainder were classified as Rest of state.
Remoteness areas are based on the Australian Bureau of Statistics Remoteness Area classification for the Statistical Area 1 (SA1) the household is located in.
The top 1% most remote SA1s in Australia were excluded from the sampling frame, and the survey has poor coverage of very remote areas. For output purposes the categories 'Remote Australia' and 'Very Remote Australia' have been combined.
The ABS Remoteness Area classification is based on the Accessibility/Remoteness Index of Australia (ARIA+) produced by the National Centre for Social Application of Geographic Information Systems (GISCA) at the University of Adelaide.
Remoteness area boundaries can be downloaded from the Australian Bureau of Statistics web site www.abs.gov.au
A shortened version of the General Functioning Subscale of the McMaster Family Assessment Device was used to classify families into four levels of functioning. This ranged from very good through to poor, with poor indicating unhealthy family functioning likely to require clinical intervention. Of all families in the survey 3.7% had a poor level of family functioning.
Boterhoven de Haan KL, Hafekost J, Lawrence D, Sawyer MG, Zubrick SR (2014). Reliability and validity of a short version of the general functioning subscale of the McMaster Family Assessment Device. Family Process. doi: 10.1111/famp.12113