METHODOLOGY

This page consists of the following topics:

• Scope
• Sample Design
• Weighting, Benchmarking and Estimation
• Perturbation of Estimates

SCOPE

The scope of the 2016 Personal Safety Survey was persons aged 18 years and over in private dwellings across Australia (excluding very remote areas). Interviews were conducted with one randomly selected person aged 18 years or over who was a usual resident of the selected household.

Private dwellings are:

• Houses
• Flats
• Home units
• Any other structures used as private places of residence at the time of the survey

Usual residents are people who usually live in a particular dwelling and regard it as their own or main home. People usually residing in non-private dwellings, such as hotels, motels, hostels, hospitals, nursing homes, or short-stay caravan parks were not in scope.

Both urban and rural areas in all States and Territories were included in the survey, except for very remote areas of Australia. The following groups were also excluded from the scope of the survey:

• Visitors at a dwelling whose usual place of residence is Australia (as they would have their chance of selection at their usual residence)
• Overseas visitors intending to stay in Australia for less than 12 months
• Non-Australian diplomats, non-Australian diplomatic staff and non-Australian members of their household
• Members of non-Australian defence forces stationed in Australia and their dependants
• People who usually reside in non-private dwellings
• Households where all residents are aged less than 18 years

SAMPLE DESIGN

The sample size, distribution and method of selection for the 2016 PSS were based on a number of factors:

• Key estimates required to be produced from the survey
• Level of disaggregation and accuracy at which these key survey estimates were required
• Costs and operational constraints of conducting the survey

The aim of the survey was to produce certain key estimates of interest with an acceptable level of quality. The sample design included the quality requirement that relative standard errors (RSEs) of less than 25% would be obtained for the following key estimates:

• experience of violence in the last 12 months
• experience of partner violence in the last 12 months, and
• experience of sexual assault in the last 12 months

Each of these key estimates were then required to be disaggregated for:

• Women: for each State and Territory (and at the national level)
• Men: at the national level. While the survey was not designed to provide State/Territory level data for men, estimates of acceptable quality are able to be produced for some of the larger States

The PSS 2016 sample was designed to meet these requirements as closely as possible whilst taking into consideration the overall costs and operational constraints of conducting the survey.

The sample for women was allocated roughly equally in each State and Territory in order to provide sufficiently reliable State and Territory and national level estimates for women. The sample for men was allocated to States and Territories roughly in proportion to their respective population size, in order to provide sufficiently reliable national level estimates for men.

In order to target the differential numbers of male and female sample, dwellings were assigned as either male (where an interview with a male aged 18 years and over was required) or female (where an interview with a female aged 18 years and over was required). One in-scope person of the pre-assigned gender was then randomly selected from each dwelling. Where the household did not contain an in-scope resident of the pre-assigned gender, an in-scope resident of the opposite gender was randomly selected, referred to as gender-selection 'flipped' household (for further information refer to Survey Development and Data Collection page of this User Guide). Due to relative differences in the number of male to female dwelling selected in the sample, a large proportion of the final male sample came from gender-selection 'flipped' households.

Lastly, response rates to the survey were expected to be impacted by a number of operational factors, designed to help ensure the safety of respondents, the safety of interviewers and also to help ensure data integrity. These included:

• The part voluntary nature of the survey
• Interviews to be conducted in a private interview setting as required
• A rule that no proxy interviews were allowed for the voluntary component
• The sensitive nature of the survey content

Due to these factors, the sample design also catered for lower response rates, expecting to attain around a 72% response rate.

In the 2016 PSS a total sample of 36,495 households were selected, comprising 29,421 female and 7,074 male pre-assigned gender households. Taking account expected sample loss and an anticipated response rate of 72%, this selected sample was designed to achieve around 22,200 fully responding households (or approximately 16,100 females and 6,100 males, incorporating the results of gender-selection 'flipped' households into these expected outcomes).

Actual numbers of fully responding households achieved are available in the Response Rates page of this User Guide.

WEIGHTING, BENCHMARKING AND ESTIMATION

WEIGHTING

Weighting is the process of adjusting results from a sample survey to infer results for the total in-scope population. To do this, a 'weight' is allocated to each sample unit corresponding to the level at which population statistics are produced. For the 2016 PSS, this is at a person level. The weight can be considered an indication of how many population units are represented by the sample unit.

Selection weights

The first step in calculating weights for each person was to assign an initial weight. The initial person weight was derived from the initial household weight (inverse of the probability of the household being selected in the survey) multiplied by the total number of in-scope males or females in the household depending on the assigned gender for the household. For example, if the probability of a household being selected in the survey was 1 in 600, and the household contained 3 in-scope females, then the selected person within the household would have an initial weight of 1,800 (ie 600 x 3). That is, the selected person in the sample represents 1,800 other persons in the population. Initial person weights took into account an increase in male-only household representation which was a result of households that were subject to gender selection flipping from their pre-assigned gender.

BENCHMARKING

Using information based on observations by interviewers at the dwelling, as well as additional information collected from non-fully responding respondents as part of the compulsory component of the survey, analysis was undertaken to ascertain whether there were any particular categories of persons that were over or under-represented in the sample. This over or under-representation in the sample can be corrected using a non-response adjustment and/or through calibrating the weights to population benchmarks. Only the calibration of weights to population benchmarks was adopted in 2016 PSS.

Benchmarks are independent estimates of the size of the population of interest. Weights are calibrated against independent population benchmarks to ensure that the survey estimates conform to the independently estimated distribution of the population, with respect to the benchmark categories, rather than to the distribution within the responding sample itself.

The PSS survey estimates were benchmarked to the estimated resident Australian population aged 18 years and over who were living in private dwellings (excluding very remote areas of Australia), simultaneously using the following benchmark categories:

Number of persons by:

1. State or territory by Capital city/balance of state by Age groups by Sex
2. State or Territory by Social marital status (Married (incl. registered or de facto) and Not married) by Sex
3. State or Territory by broad Country of birth (Australia, Main English Speaking categories and Other) by Sex
4. State or Territory by Labour force status (Full Time Employed, Part Time Employed, Unemployed, and Not In the Labour Force) by Sex
5. Age group (slightly more detailed) by Sex

Two benchmark categories (1 and 5) were benchmarked to the estimated resident population living in private dwellings in non-very remote areas of Australia as at February 2017 based on the 2016 Census of Population and Housing. The in-scope Estimated Resident Population was estimated to be 18,401,503 as at February 2017. The benchmarks, and hence the estimates from the survey, do not (and are not intended to) match estimates of the total Australian estimated resident population (which include persons living in very remote areas of Australia and persons in non-private dwellings, such as hotels) obtained from other sources.

The remaining benchmark categories (2,3,4) were based on survey estimated benchmarks. Benchmarks are considered to be survey estimated if they are obtained from a sample survey and as such, have a non-negligible level of sample error associated with them. The monthly Labour Force Survey (as per February 2017, prior to any revised seasonal adjustments which were released in April 2017) provided the survey estimated benchmarks for labour force status, social marital status and broad country of birth categories. The survey estimated benchmarks were aligned to the in-scope resident population aged 18 years or more, who were living in private dwellings in each state and territory (excluding very remote areas of Australia), as at February 2017. They were also made to represent the same population as the demographic benchmarks with respect to state/territory (although note difference below for males), part of state, age group and sex. The sample error associated with these survey estimated benchmarks was incorporated into the standard error estimation.

The 'state' population benchmark for benchmark 1, for females, consisted of all six states and two territories. Males were not benchmarked to the state/territory component of this benchmark.

For the survey estimated benchmarks 2, 3 and 4, Tasmania, Northern Territory and Australian Capital Territory were collapsed for both males and females.

For benchmark 1 and 5 'Age group' was benchmarked against 5 year age groups, between the 18-19 years and 80 years and over groups. In addition, for benchmark 1, grouping of 18-19 and 20-24 occurred for females living in Tasmania balance of state.

Note for male estimates: Tasmania, Northern Territory and Australian Capital Territory have not been benchmarked to their State/Territory male populations and therefore contribute to national estimates only. Based on the contributing benchmarks to the weights, male estimates have been benchmarked to total male populations in each of New South Wales, Victoria, Queensland, South Australia and Western Australia. However, as outlined in the Sample Design section of this page, the sample was not designed to produce estimates for states/territories for males and therefore detailed estimates are likely to have associated high standard errors (for more details on standard errors, see Data Quality and Technical Notes page of this User Guide).

ESTIMATION

Estimation is a technique used to produce information about a population of interest, based on a sample of units (i.e. persons) from that population. Each record in the 2016 PSS has a person weight. Information for sampled persons is multiplied by the weights to produce estimates for the whole population (or the population of interest).

If each person's weight were to be ignored when analysing the data to draw inferences about the population, then no account would be taken of each person's differing chance of selection or of different response rates across population groups, with the result that the estimates produced could be seriously biased. The application of weights ensures that estimates will conform to an independently estimated distribution of the population by certain categories including state, age and sex rather than to the distributions within the sample itself.

Replicate weights have also been included - 60 person replicate weights. The purpose of these replicate weights is to enable calculation of the relative standard error (RSE) for each estimate produced from the survey. Further information on replicate weights is provided in the Data Quality and Technical Notes page of this User Guide.

Users should take into consideration the quality of the estimates when interpreting data from the PSS. For further information about how to calculate RSE's and conduct significance testing, refer to the Data Quality and Technical Notes page of this User Guide. For other related information on other factors to consider when interpreting results, refer to the individual topic pages in this User Guide.

PERTURBATION OF ESTIMATES

To minimise the risk of identifying individuals in aggregate statistics, a technique called perturbation is used to randomly adjust cell values. Perturbation involves a small random adjustment of the statistics and is considered the most satisfactory technique for avoiding the release of information that could identify individual survey respondents while maximising the range of information that can be released. These adjustments have a negligible impact on the underlying pattern of the statistics.

After perturbation, a given published cell value will be consistent across all tables. However, adding up cell values to derive a total will not necessarily give the same result as the published totals. As such, proportions may add to more or less than 100%. Users are advised to use the published totals rather than deriving totals based on the component cells.

Cells with relatively small values may be proportionally more affected by perturbation than large values. Users are advised against conducting analyses and drawing conclusions based on small values.

All data presented in the Personal Safety Survey, Australia 2016 (cat. no. 4906.0) have had perturbation applied to estimates. The introduction of perturbation in publications ensures that published statistics are consistent with statistics released via services such as TableBuilder. It should be noted that this method of perturbation is not possible to apply to the Detailed Microdata product, and statistics from this product may not be consistent with published or TableBuilder statistics.