Australia

State

Household; Individual

All members of private dwellings in Australia, with the following exceptions:

• certain diplomatic personnel of overseas governments, customarily excluded from censuses and surveys;
• overseas residents in Australia (that is, persons who had stayed or intended to stay in Australia less than one year);
• members of non-Australian defence forces (and their dependents) stationed in Australia;
• residents of institutions (such as hospitals and other health care institutions, military and police installations, correctional and penal institutions, convents and monasteries);
• people living in remote and sparsely populated areas; and
• residents of dwellings that were not primary places of residence.

The sampling unit is the household, and members of those households will be traced over an indefinite life. The main sample was automatically extended over time by following rules that add to the sample:

• any children born to or adopted by members of the selected households;
• new household members resulting from changes in the composition of the original households;
• a new household member that arrived in Australia for the first time after 2001 (or after 2011 once the top-up sample was added).

Longitudinal/panel/cohort study

Annual

Multi-stage stratified random sample. The sampling unit is the household, and members of those households have been and will be traced over an indefinite life. The households were selected using a multi-staged approach. First, a sample of 488 Census Collection Districts (CDs) were selected from across Australia (each of which consists of approximately 200 to 250 households). Second, within each of these CDs, a sample of 22 to 34 dwellings was selected, depending on the expected response and occupancy rates of the area. The selections were made after all dwellings within each of the CDs were fully listed. Finally, within each dwelling, up to three households were selected to be part of the sample. In order to retain cross-sectional representativeness in the sample an additional 2,153 households were added to the sample as part of a general top-up in wave 11. The top-up sample was selected using the same methodology as the original HILDA sample.

The vast majority of the data were collected though face-to-face interviews. While telephone interviews and assisted interviews were conducted to ensure a high response rate, they are only used as a last resort. The percentage of respondents interviewed by telephone was between 0.5% in Wave 1 and 10.1% in Wave 8 (Main Sample) and between 1.8% in Wave 11 and 6.0% in Wave 15 (Top-Up Sample). In Wave 20, due to the COVID-19 pandemic and associated restrictions and social distancing requirements, 95.7% of the interviews were conducted by telephone. Self-completion mode was also used for data collection since all persons completing a person questionnaire were also asked to complete Self-Completion Questionnaire which the interviewer collects at a later date or is returned by mail. In Waves 1-19 a paper Self-Completion Questionnaire was used but in Wave 20 an online version was also introduced. In Wave 20, 82.0% of the Self-Completion Questionnaires were completed online.

In wave 1, the selected households were sent a primary approach letter and a brochure approximately one week prior to when the interviewer was scheduled to make contact with the household. This pre-interview material marketed the survey to respondents as a study about ‘Living in Australia’ and, among other things, emphasised that participation was voluntary and provided a means for sample members to opt out of the survey prior to an interviewer calling. From wave 2 onwards, a primary approach letter and newsletter were sent to the last known address of the households approximately one month prior to when the interviewer was scheduled to make contact with the household. The newsletter provides respondents with some results from prior waves. In addition to the posted pre-interview material, households with people who had not been part of the household in the previous wave were given a New Entrants Brochure. This brochure provided more information about the purpose of the study, why they had been asked to participate, and a method to opt out of the study if they chose to. A follow-up newsletter was introduced from wave 3 onwards. In wave 1, the fieldwork was conducted in two stages. The first stage involved the interviewer working in an area over a three-week period. They visited each selected household according to the specified call-back pattern. This achieved approximately 65 per cent of the interviews from each area. The remainder of each workload was then consolidated into intensive follow-up workloads and reassigned to the most experienced interviewers. They again visited each of these households according to the specified call-back pattern. These interviewers obtained the remaining 35 per cent of the interviews from each area. From wave 2 onwards, a tracking component is incorporated into the fieldwork, splitting it into three distinct periods. All households were issued into the field for the first period, and where all the interviews had not been completed, they were reissued into the field in the next period. If a household could not be found at either one of these stages, they were put into tracking and once found were issued back into the current period if found quickly or more generally into a later period. The third period was used to finalise households that had to be traced and could not be immediately issued back into the field and also to contact some households where it was deemed beneficial to contact them in this period (for example, a household member may have been away from the household at earlier contacts or they may have been temporarily unwell or busy). Average time (minutes) taken to complete questionnaires was approximately 4-6 minutes for Household form, 4.7-17.4 minutes for Household Questionnaire, 34.4 minutes for Person Questionnaire (Wave 1 only), 28.8-40.0 minutes for Continuing Person Questionnaire, 34.2-48.0 minutes for New Person Questionnaire and approximately 20-30 minutes for Self-Completion Questionnaire.

The initial (or design) weights are derived from the probability of selecting the households into the sample. These household weights are adjusted according to information collected about all selected households (both responding and non-responding) and then calibrated so that weighted household estimates from the HILDA Survey match several known household-level benchmarks. The person-level weights are based on the household-level weights, with adjustments made based on information collected about all the people listed in the responding households. These weights are also calibrated to ensure that the weighted person estimates match several known person-level benchmarks. The cross-sectional weights integrate the main sample and the top-up sample. In both samples the initial weights are calculated as described above, followed by a non-response adjustment. Longitudinal weights were provided for the balanced panel of responding persons, Self-Completion Questionnaire respondents or enumerated persons from every wave to every other wave and for the balanced panel of any combination of a pair of waves. These weights adjust for attrition from the initial wave and are benchmarked back to the key characteristics of the initial wave.

The Wave-on-wave response rate (including fully and partially responding households) ranges from 86.9 per cent in Wave 2 to 97.0 per cent in Waves 15 and 16 for the Main Sample and from 92.3 per cent in Wave 12 to 96.2 per cent in Wave 19 for the Top-Up Sample.