The 2024 TUS was conducted by the Australian Bureau of Statistics (ABS) between August and October 2024. Previous surveys were conducted in 1992, 1997, 2006 and 2020-21.
TUS provides insight into how Australians aged 15 years and over, spend their time across a day, including:
The ABS regularly updates survey content and data collection approaches to ensure the survey remains relevant, that respondent imposition is reduced as much as possible, and that the best methods are applied within budget constraints.
For the 2024 TUS, improvements to survey content and data collection were informed by feedback from 2020–21 TUS respondents (members of the Australian community who completed the survey), an ABS survey evaluation project, and ongoing efforts to modernise ABS surveys to meet the community’s digital expectations.
2020-21 TUS respondents provided clear and consistent feedback that the survey – particularly the requirements associated with diary approach – required too much detail, were time consuming to complete and some other aspects of the survey were difficult to understand and complete.
The ABS 2020-21 TUS evaluation project considered survey response rates and response completeness (i.e., how fully a respondent has answered the questions they were asked). ABS analysis showed that while the main activity was well reported, there were significant missingness for the remainder of the diary data items such as secondary activities and contextual information such as location, use of devices and who the activity was “with” and “for”.
Regarding the ongoing modernisation of surveys, the ABS evolves its collection approach, wherever possible, to improve efficiency and reduce cost, reduce demands on the Australian public to provide their data, improve respondent experiences and take advantage of methodological and technological innovations. While these innovations are not always visible to data users, they are critical to maintaining the quality of official statistics.
In response to 2020-21 TUS respondent feedback, the ABS evaluation project, and the ongoing modernisation of ABS surveys to meet the community’s digital expectations, the ABS:
The combination of the above improvements led to a significant reduction in the average time taken to complete the digital diary – from 56 minutes in 2020-21 to 36 minutes in 2024. In addition, the average length of the household questionnaire component reduced from 35 to 19 minutes.
These improvements also delivered improvements in the quality of the 2024 TUS results, such as a reduction in missing diary data. In 2020-21 the average 24 hour diary included approximately 60 minutes of missing activity. In 2024, this average reduced to just 8 minutes per day.
While the changes above have demonstrated improvements in relation to the completeness of the TUS diary data, the impact of these changes needs to be considered when comparing 2020-21 and 2024 results. If you choose to compare across surveys, the ABS recommends using the following caveat: The 2024 Time Use Survey estimates are not fully comparable with previous collections due to changes in survey methodology.
The scope of the survey included:
The following people were excluded:
Households were randomly selected to participate in the survey. The sample was designed to support national level estimates.
The survey enumeration for 2024 was conducted in the following time periods:
The survey was collected via computer assisted web interview (CAWI) and computer assisted telephone interview (CATI). Data collection occurred in two phases:
Phase 1: An online household questionnaire was completed by any person in the household aged 15 years or over. This person provided responses to demographic and socio-economic questions on behalf of all in scope household members.
Phase 2: An online activity diary was completed by all in scope household members aged 15 years and over for two consecutive days assigned to the household. Respondents provided information about the type of activity they were undertaking and the duration of that activity. The diary also included questions about simultaneous caring. At the completion of the diary there were questions about health, feeling rushed for time and participation in unpaid voluntary work.
A small number of questionnaires and diaries were completed over the telephone for respondents who could not or chose not to complete the survey and/or diary online. For these respondents, only one diary day was collected.
Respondents completed the diary by selecting what they were doing from a pre-defined list of activities, beginning at midnight on day 1. The finish time of the first activities became the start time of the next activity and so on for the full 24 hours. Activities could be recorded based on five minute increments. This was repeated for day 2.
If an activity was not available on the list, respondents could add the activity in an ‘Activity (please specify)’ text box. These activities were office coded to the appropriate activity.
Only one activity was captured at a time, and respondents were instructed to select what they considered was their main activity.
Rigorous quality assurance was applied throughout the survey process.
As discussed above, the survey questionnaire and diary were enhanced from the 2020-21 version, incorporating updates as well as feedback from users. The new diary was tested with a diverse range of respondents prior to the commencement of the survey to ensure clarity, relevance, and ease of completion.
The sample was designed specifically for online collection, taking into account:
It was estimated that to achieve the required sample of 10,000 diary days 25,912 dwellings needed to be included in the sample.
Information was collected from 6,899 households. From these households, 10,613 persons provided a total of 19,319 diary days.
The achieved sample was higher than the expected 10,000 diary days. This was attributed to several factors:
All selected households received a postal invitation with instructions to complete the survey online. Those unable or unwilling to participate online were offered the opportunity to complete the survey via telephone. There was no face-to-face follow-up.
Two reminder letters were sent to non-responding households. Due to the absence of face-to-face follow-up, sample selection and follow-up procedures were designed to minimise bias and optimise representativeness.
An instructional video and written instructions were available on the ABS web site to help with diary completion.
Survey progress was monitored throughout the enumeration period. Households that started but did not complete the survey online were followed up with reminder emails.
Respondents were sent reminder emails and text messages before, during and after their allocated diary days.
There were minimal processing and coding of diary data:
Quality assessment thresholds and some minor edits were applied to the diaries. Only diaries that met the quality threshold were retained for the final data file. For example:
Respondents were asked to use their best judgement in determining their main activity at any point in time. This could lead to underreporting of activities not consistently considered to be a main activity (eg. Listening to music).
As only a sample of people were surveyed on certain days, results needed to be converted into estimates for the whole population. This was done through a process called weighting.
The person weights were then calibrated to align with independent estimates of the in-scope population, referred to as ‘benchmarks’. The benchmarks used additional information about the population to ensure that:
Estimates from the survey were obtained by weighting the diary day responses to represent the in-scope population of the survey.
Benchmarks align to the Estimated Resident Population (ERP) aged 15 years and over as at August 2024 (after exclusion of people living in non-private dwellings, very remote areas of Australia, and discrete Aboriginal and Torres Strait Islander communities).
The weighted data was compared to the previous survey and other ABS data sources to make sure estimates were in line with expectations.
Estimates from 2024 TUS may differ from similar data items produced from other ABS collections for several reasons. Differences in sampling errors, scope, collection methodologies, reference periods, seasonal and non-seasonal events may all impact estimates.
The 2024 TUS is designed to provide rich insight into how various population subgroups report spending their time. It is not designed to provide measures of prevalence (such as employment, voluntary work, disability status, or caring status).
The proportion who participated in an activity does not reflect the prevalence rate of a characteristic in the general population. For example, the proportion of those who participated in ’caring and support’ activities is not equivalent to the proportion of carers in the population. This is because participants complete only two diary days and a carer may not have provided any care on those diary days.
The Activity Classification for 2024 TUS has been updated based on findings from the 2020-21 survey, improvements to the collection, and clearer labelling.
The detailed Activity Classification is available in the Data Item list in the Data downloads section.
Activities are grouped into four overarching types of time:
A person who reported in their time use diary that they spent at least five minutes in their day doing an activity.
This is the proportion of people in a population who spent at least five minutes on an activity in a day. This is calculated:
\(\mathsf{\large{\frac{\text{Persons in population who participated in activity}}{\text{Total persons in population}}\times 100}}\)
For example, the proportion of females who participated in domestic activities:
\(\mathsf{\large{\frac{\text{Females who participated in domestic activities}}{\text{Total female population}}\times 100}}\)
Adding proportions of participants across multiple activities can lead to double counting because some people may be included in more than one category. As a result, the total will not be accurate.
The average time spent on an activity by people who reported spending at least five minutes in the day doing this activity, is calculated:
\(\mathsf{\large{\frac{\text{Total time spent on activity in a day by persons in the population}}{\text{Persons in population who participated in activity}}}}\)
For example, the average time spent per day by females who participated in domestic activities:
\(\mathsf{\large{\frac{\text{Total time females spent on domestic activities in a day}}{\text{Females who participated in domestic activities}}}}\)
This average is particularly useful for reporting the average time spent on activities that the whole population did not participate in, for example work or looking after children. This is because it only includes people who actually did the activity. For example, among those who worked, the average time spent working was 7 hours and 58 minutes per day.
Summing the ‘average time spent per day, of persons who participated in activity’ for more than one activity will double count people who appear in more than one category and will not give an accurate total.
The average time spent per day of persons who participated in separate activity categories cannot be summed together to calculate a total average.
The average time spent on an activity by all people, regardless of whether they reported doing that activity in their day or not, is calculated:
\(\mathsf{\large{\frac{\text{Total time spent on activity in a day by persons in the population}}{\text{Total persons in population}}}}\)
For example, the average time spent per day on domestic activities by all females:
\(\mathsf{\large{\frac{\text{Total time females spent on domestic activities in a day}}{\text{Total female population}}}}\)
When only a small portion of the population participates in an activity, the average time spent per day across the whole population will be very low. In such cases, this average may not be very useful. For example, the average time spent per day on looking after children of the total population will include people who do not spend any time caring for children.
The average time spent per day of the total population from separate activity categories can be summed together to calculate a total average.
In a small number of instances there are inconsistencies between the data provided in Phase 1 for the household questionnaire and Phase 2 for the diary. For example, a person may be reported as unemployed in the household questionnaire but then reports spending time on employment in their diary.
Differences could be due to:
Two types of error are possible when producing estimates from a sample-based survey:
Refers to factors unrelated to sample selection which produce data results that don’t accurately represent the population. They can occur at any stage throughout the survey process. For example:
Refers to the expected difference between survey estimates and the true value that would result from surveying the whole population. It is the result of random variation and can be estimated using measures of data variance.
One measure of sampling error is the standard error (SE). There are about two chances in three that an estimate will differ by less than one SE from the figure that would have been obtained if the whole population had been surveyed. There are about 19 chances in 20 that an estimate will differ by less than two SEs.
The relative standard error (RSE) is a useful measure of sampling error. It is the SE expressed as a percentage of the estimate:
\(\mathsf{RSE\%=\frac{SE}{estimate}\times100}\)
Only estimates with RSEs below 25% are considered reliable for most purposes. Those with RSEs between 25% and less than 50% are included in the publication but flagged to indicate high associated SEs and should be used with caution.
Estimates with RSEs of 50% or more are also flagged and considered unreliable for most purposes. Their RSE values are not published.
Another measure of sampling error is the margin of error (MOE). This describes the distance from the population value that the sample estimate is likely to be within and is particularly useful to understand the accuracy of proportion estimates. It is specified at a given level of confidence. Confidence levels typically used are 90%, 95% and 99%.
For example, at the 95% confidence level, the MOE indicates that there are about 19 chances in 20 that the estimate will differ by less than the specified MOE from the population value (the figure obtained if the whole population had been surveyed). The 95% MOE is calculated as 1.96 multiplied by the SE:
\(\mathsf{MOE=SE\times1.96}\)
The RSE can also be used to directly calculate a 95% MOE by:
\(\mathsf{MOE(y)\approx \frac{RSE(y)\times y}{100}\times 1.96}\)
The MOEs in this publication are calculated at the 95% confidence level. This can easily be converted to a 90% confidence level by multiplying the MOE by:
\(\mathsf{\large{\frac{1.615}{1.96}}}\)
or to a 99% confidence level by multiplying the MOE by:
\(\mathsf{\large{\frac{2.576}{1.96}}}\)
Depending on how the estimate is used, an MOE of greater than 10% may be considered too large to inform decisions. For example, a proportion of 15% with an MOE of plus or minus 11% means the estimate could lie anywhere between 4% to 26%. It is important to consider ranges when using estimates to make assertions about populations.
A confidence interval expresses the sampling error as a range in which the population value is expected to lie at a given level of confidence. A confidence interval is calculated by taking the estimate plus or minus the MOE of that estimate. In other words, the 95% confidence interval is the estimate +/- MOE.
Proportions or percentages formed from the ratio of two estimates are also subject to sampling errors. Error size depends on the accuracy of the numerator and denominator. A formula to approximate the RSE of a proportion is given below. This formula is only valid when the numerator \(\mathsf{\small{(x)}}\) is a subset of the denominator \(\mathsf{\small{(y)}}\):
\(\mathsf{RSE (\frac {x}{y}) = \sqrt {[RSE(x)]^2 - [RSE(y)]^2}}\)
When calculating measures of error, it may be useful to convert RSE or MOE to SE. This allows the use of standard formulas involving the SE. The SE can be obtained from RSE or MOE using the following formulas:
\(\mathsf{\large{SE = \frac{RSE\% \times estimate}{100}}}\)
\(\mathsf{\large{SE = \frac{MOE}{1.96}}}\)
Differences between survey estimates – be they counts or percentages - can be calculated from published data, but are also subject to sampling error. The size of this error depends on the standard errors (SEs) of the individual estimates and the statistical relationship (correlation) between them.
An approximate SE of the difference between two estimates \(\mathsf{\small{(x-y)}}\) may be calculated by the following formula:
\(\mathsf{SE(x-y) \approx \sqrt {[SE(x)]^2 + [SE(y)]^2}}\)
While this formula will only be exact for differences between unrelated characteristics or sub-populations, it provides a reasonable approximation for differences likely to be of interest in this publication.
When comparing estimates across surveys or between populations within a survey, it's important to assess whether differences are real or simply due to sampling variability.
One way to do this is by testing for statistical significance. This involves calculating the standard error of the difference between two estimates \(\mathsf{\small{(x \text{ and } y)}}\) and using it to compute a test statistic, as shown in the formula below:
\(\mathsf{\Large{\frac{[x-y]}{SE(x-y)}}}\)
Where
\(\mathsf{{SE(y) \approx \frac{RSE(y) \times y}{100}}}\)
A test statistic greater than 1.96 provides strong evidence of a statistically significant difference between the two populations at the 95% confidence level. If the value is 1.96 or lower, the difference may be due to sampling variability, and we cannot confidently conclude that a real difference exists.
The 2024 TUS release presents national estimates.
Data cubes for this release present tables showing proportions and average time spent on activities, along with their corresponding MOEs and RSEs. A data item list is also available in the Data downloads section.
Detailed microdata for 2024 TUS is available in DataLab for users who want to undertake in-depth analysis using a range of statistical software packages in the secure ABS environment. Refer to Microdata: Time Use.
Customized statistical tables are available upon request on a cost-recovery basis. However, confidentiality requirements and sampling variability may limit what data can be provided.
To inform requests, refer to the Data Item List in the Data downloads section of the ‘How Australians use their time’ publication.
If you would like to request a consultancy, complete the Consultancy request form.
For more general information, get in touch via our Contact us page.
The Census and Statistics Act 1905 authorises the ABS to collect statistical information and protects information from being published in a way that could identify persons or organisations. The ABS therefore ensures that information about individual respondents cannot be derived from published data.