SAMPLE DESIGN

SAMPLE SELECTION

The LFS sample consists of three components, which are separately identified and sampled:

• private dwellings (houses, flats, etc.);
• non-private dwellings (hotels, motels, hospitals, homes for the aged, university colleges, boarding houses, predominantly non-long stay caravan parks etc.); and
• dwellings in discrete Aboriginal and Torres Strait Islander communities.

The sample of private dwellings is obtained by a multi-stage approach. Using the SA4 structure of the Australian Statistical Geography Standard (ASGS), Australia is first divided into 88 geographical areas. In general, these areas are then sub-divided and grouped to form a strata. This strata is based on locality, population density, remoteness and projected population growth.

Within these strata, a multi-stage sample of geographic areas is selected, where the area sampling units are derived hierarchically (as described below). The areas at each stage are aggregations of mesh blocks. Meshblocks are the finest geographic unit in the ASGS and usually contain between 30 and 60 dwellings. The area sampling units are derived as follows:

• in the first stage of selection, the area sampling units are large enough to contain sufficient dwellings to provide enough sample for the LFS for five years. A sample of these areas is randomly selected by systematic sampling (with the probability of selection proportional to number of dwellings) to represent each stratum;
• for the second stage of selection, the sampling unit is a region contained within the selected first stage area units. One of these is randomly selected with a probability proportional to the number of dwellings. These area units are an aggregation of a number of mesh blocks to be large enough to provide at least five separate samples of dwellings for the third stage;
• in the third stage, a sample of dwellings in the selected block is taken using systematic equal probability sampling.

In less populated areas, there is an additional stage of selection in which selected first stage area units are grouped geographically, and selection occurs within these groups. This ensures the sample is not too geographically spread and is not prohibitively expensive to enumerate.

The sample of non-private dwellings is selected from a list of non-private dwellings in Australia. This is compiled from the 2011 Census of Population and Housing and updated using the Geocoded National Address File (G-NAF). A sample is taken from this list that represents each geographical area and different types of dwellings. For smaller non-private dwellings, each occupant is included in the survey; while for larger dwellings a sub-sample of occupants is taken.

The sample of dwellings in areas primarily consisting of discrete Aboriginal and Torres Strait Islander communities is obtained by:

• identifying those states and territories with sufficient population in discrete Aboriginal and Torres Strait Islander communities such that a separate sample can be selected from areas containing Aboriginal and Torres Strait Islander communities;
• grouping mesh blocks which contain Aboriginal and Torres Strait Islander communities that have community and outstation links and selecting a sample from these groups with a probability proportional to the number of dwellings in the group, and;
• taking a sample of dwellings in those selected communities using systematic equal probability sampling.

ALLOCATION OF SAMPLE

The LFS is primarily designed to provide reliable estimates of the key labour force statistics for the whole of Australia. The secondary design is to produce key LFS statistics for each state and territory. The most accurate national estimates would be obtained if the total sample across Australia was allocated in proportion to the population of each state or territory. Conversely, to provide estimates of similar accuracy for each state or territory, approximately equal sample sizes would be needed for each.

The actual allocation of LFS sample across the states and territories is designed to balance the accuracy of national estimates with state and territory estimates. The proportion of the population in the sample (known as the sampling fraction) is larger in the states and territories with smaller populations, but not to the extent that would realise identical sample sizes for each state and territory. Within each state or territory, each dwelling has the same probability of selection.

SAMPLE ROTATION

One of the primary requirements of the LFS is to provide a measure of change in the characteristics of the labour force over time, for both month-to-month variation and longer trends over time. The best estimates of change from one month to the next would be obtained if the survey was collected from the same sample of dwellings each month while providing for population growth. However, it is neither reasonable nor representative to continually retain the same respondents in the survey. Instead a proportion of the sample is replaced each month in order to measure changes in the labour force while also ensuring the survey is representative. This procedure is known as sample rotation.

Since the monthly LFS began in 1978, one-eighth of the sample has generally been replaced each month. The sample can be thought of as consisting of eight sub-samples (or rotation groups), and each month one rotation group which is a new sample of dwellings replaces a sample of dwellings which had been in the LFS for the previous eight months. The dwellings in the replacement sample usually come from the same areas as the dwellings they replaced.

Sample rotation enables reliable measures of monthly change in labour force statistics to be compiled, as seven-eighths of the sample from one month are retained for the next month's survey. At the same time, the sample rotation procedure ensures that no dwelling is retained in the sample for more than eight months.

The component of the sample that is common from one month to the next makes it possible to match the characteristics of most of the people in those dwellings: this group is referred to as the 'matched sample'. The availability of this matched sample permits the production of estimates of 'gross flows' - the number of people who change labour force status between successive months.