Microdata: Time Use

Describes the definitions, concepts, methodology and estimation procedures used in the Time Use Survey

Introduction

Overview

This publication contains details about the 2006 Time Use Survey (TUS). It includes information about the survey objectives, the development process, content of the survey and the methods and procedures used in the collection and processing of data. It also includes information about the quality and interpretation of the survey results and about the products and services available. 

The 2006 Time Use Survey was conducted by the Australian Bureau of Statistics (ABS) to obtain information about the way people allocate time to different kinds of activities. It was conducted over four periods during 2006 in order to balance seasonal influences which affect time use patterns. The first national Time Use Survey was conducted by the ABS in 1992 which followed from a pilot test in Sydney in 1987. The Time Use Survey was conducted again in 1997. Comparison of the national data can be made across time. Time use surveys collect information about all activities people engage in during a specified period. As a result of this, the range of information they provide is very broad. 

Information was collected from households using an interview and a diary. The interview was designed to provide characteristics by which the population can be disaggregated into sub-groups in order to examine differences in time use. The diary contained information about the way respondents spent their time over a two day period. 

Details of the 2006 Time Use Survey were tabled in Federal Parliament in accordance with section 6(3) of the Australian Bureau of Statistics Act 1975 and the survey was conducted under the authority of the Census and Statistics Act 1905. The ABS sought the willing cooperation of people living in private dwellings. Under its legislation the ABS cannot release identifiable information about households, families or individuals. The confidentiality of all information provided by respondents is guaranteed.

Survey background and purpose

Overview

The Australian time use surveys have been supported by submissions from Commonwealth and State government and non-government organisations for policy development and monitoring in areas as diverse as community services, employment, women's affairs, families, education, communication, health, town and traffic planning, media and leisure. Time allocation data can be analysed in relation to the demographic, socio-economic and other personal characteristics collected in the survey.

Objectives of the 2006 survey

The 2006 Time Use Survey was designed to identify 

  • unpaid work in the household and the breakdown of this work between men and women and by life stage;
  • daily life patterns and support needs of various groups such as older persons, unemployed persons and persons with disabilities;
  • labour force issues such as the comparison of the hours of paid work for men and women and barriers to work force participation;
  • time spent on voluntary work, caring and other unpaid community work;
  • patterns of leisure activity;
  • transport issues;
  • the production of services not included in estimates of Gross Domestic Product (GDP); and
  • to make comparisons with the 1992 and 1997 surveys to identify changes in patterns of time use.

The survey also provides information on:

  • balancing paid work with other aspects of life;
  • caring for people with disabilities and frail older people;
  • caring for children;
  • community participation;
  • fitness and health activities;
  • travel;
  • use of technology; and
  • outsourcing of domestic tasks.

Time use as a social indicator

Although each person has 24 hours in a day, the demands on time vary greatly from person to person. While personal choices contribute to differences in time use, life-cycle stage, family commitments, sex and other socio-economic characteristics have determining effects. Time use could be a useful indicator of well-being, having implications for income, health, equality of access to opportunities, and personal fulfilment. Comparisons can be made between various sub-groups, between a sub-group and the population as a whole, and between Australia and other countries. As this is the third national time use survey, comparisons can also be made over time. 

Having too much to do, or too little to do, can both be seen as conditions of disadvantage, affecting income, health and morale. Persons living alone, for instance, who report no social contact and considerable stretches of time doing 'nothing' may not see themselves as fortunate. An unemployed person may prefer income-producing work to free time. However, having too much to do in some areas can interfere with adequate access to other uses of time. Someone who is caring for an elderly parent with a disability, for instance, may not have enough unencumbered time for income earning activities, with implications for that person's financial security in the future. Another person who works for income and also carries the main responsibility for parenting and housework may not have adequate leisure or rest time to maintain good health. The 2006 Time Use Survey explores the concept of 'time stress'. Questions in the diary record people's perception of how often they have too much or too little time and the reasons for this as well as their satisfaction with the way they spend their time.

Unpaid work

The well-being of many people depends on unpaid work undertaken by individuals. If services such as: cooking; cleaning; domestic management; home and car maintenance; care of the frail, sick and those with disabilities; and care of children are not provided 'free', they have to be paid for as a market transaction. Therefore, some measurement of unpaid or non-market work, along with measurements of paid work and production, are necessary for a comprehensive picture of national production and consumption. 

The ABS uses measurements of time spent in unpaid work as one of the statistical bases for estimating the value of unpaid work in Australia within a national accounting framework. Given the absence of an agreed international standard definition, unpaid housework forms part of the concept of committed time. The concept of committed time is based on a restricted version of the third person criterion - the market replacement criterion where an unpaid activity is considered to be unpaid work if the output produced can be purchased in the market or if the activity can be delegated in exchange for payment. The major activity groups that will be included in committed time are broadly comparable with the classification of unpaid work adopted by the ABS in Occasional Paper: Measuring Unpaid Household Work, 1992 (Cat. no. 5240.0). 

The ABS restructured the Time Use Activity Classification within the framework of Dagfinn Ås' four way division of time (Ås 1982) prior to the 1997 survey. For a discussion of the identification and measurement of unpaid work and the four way division of time, see Time Use Survey, Australia Users' Guide, 1997 (Cat. no. 4150.0).

Activity classification

The activities that people record in their diaries are classified according to the Time Use Activity Classification, which was redeveloped for the 1997 Time Use Survey to align with the four types of time: 

  1. Necessary time; includes activities which serve basic physiological needs such as sleeping, eating, personal care, health and hygiene.
  2. Contracted time; includes paid work and regular education. Activities within this category have explicit contracts which control the periods of time in which they are performed. These activities, therefore, constrain the distribution of other activities over the rest of the day.
  3. Committed time; describes activities to which a person has committed him/herself because of previous acts or behaviours or community participation such as having children, setting up a household or doing voluntary work. The consequent housework, care of children, shopping or provision of help to others are committed activities. In most cases, services could be bought to provide the same activity (e.g. an exchange could be made of time for money). The unpaid work activities which are identified in the satellite national accounts are all committed time activities.
  4. Free time; is the amount of time left when the previous three types of time have been taken out of a person's day. Many free time activities are considered as leisure, but not all. Leisure time is subjective and depends on a particular person's point of view. In fact, many activities included in committed time can be considered to be leisure time activities by some people (e.g. gardening, furniture making). The only way to obtain more free time is for contracts and commitments to be changed or to spend less time on necessary time activities (e.g. sleep less).

The 2006 Time Use Activity Classification used nine major activity groups, arranged to relate to the above typology:

Necessary time
  1. Personal care activities
Contracted time
  1. Employment activities
  2. Education activities
Committed time
  1. Domestic activities
  2. Child care activities
  3. Purchasing activities
  4. Voluntary work and care activities
Free time
  1. Social and community interaction
  2. Recreation and leisure

A more detailed list of activity categories can be found in Appendix 1.

Survey methodology

Scope and coverage

Scope of the survey

The scope of the estimates from this survey is all usual residents in private dwellings throughout Australia, excluding very remote dwellings. The survey collected information by personal interview from usual residents of private dwellings in urban and rural areas of Australia, covering about 98 per cent of the people living in Australia. Private dwellings are houses, flats, home units, caravans, garages, tents and other structures that are used as places of residence at the time of interview. Long-stay caravan parks are also included. These are distinct from non-private dwellings which include hotels, boarding schools, boarding houses and institutions. Residents of non-private dwellings are excluded. 

The survey excludes: 

  • households which contain members of non-Australian defence forces stationed in Australia;
  • households which contain diplomatic personnel of overseas governments; and
  • households in collection districts defined as very remote or Indigenous Communities.

Also excluded were persons living in very remote parts of Australia. The exclusion of these persons will have only a minor impact on any aggregate estimates that are produced, particularly as the Time Use Survey is designed to produce National rather than State/Territory estimates.

Coverage rules

Information was collected from usual residents only. Usual residents were residents who regarded the dwelling as their own or main home. Others present were considered to be visitors and were not asked to participate in the survey.

Sample design and selection procedures

Sample design

For the Time Use Survey, a target sample size of 3,870 households throughout all States and Territories was sufficient to provide: 

  • detailed person-level information for Australia;
  • detailed household-level information for Australia;
  • relatively detailed person day-level information for Australia for weekend/Saturday/Sunday day types;
  • relatively detailed data for Capital City/Balance of State or Territory; and
  • estimates for those characteristics which are relatively common and for sub-populations which are relatively large and spread fairly evenly geographically.
Sample selection

The Time Use Survey had special requirements and constraints. Time use may vary according to the day of the week and particularly between weekdays, Saturdays and Sundays, and the survey was designed to account for this. In the 1997 survey all days of the week were surveyed in equal proportion and estimates were produced for weekdays and weekends. For the 2006 survey weekend data was to be split into Saturday and Sunday and estimates produced for weekdays, Saturdays and Sundays. In order to achieve acceptable standard errors for estimates produced for Saturdays and Sundays, twice as many Saturdays and Sundays were surveyed as every weekday. Therefore, approximately 1/9 of the sample was assigned to each weekday and 2/9 of the sample was assigned to each of Saturday and Sunday.

The days of the week were surveyed in approximately these proportions in each of the four collection periods during the year, with school and public holidays represented in approximately the same proportion as they occurred during the year. Diaries were therefore to be completed on specified days. 

Collection periods were as follows: 

  • 20th February - 4th March 2006;
  • 24th April - 6th May 2006;
  • 26th June - 8th July 2006;
  • 23rd October - 4th November 2006.

The survey was conducted using a stratified multi-stage area sample of private dwellings (houses, flats, etc.) in both urban and rural areas in all States and Territories, except for very remote parts of Australia. Usual residents of the dwelling aged 15 years and over were asked to participate in the survey.

The 2006 Time Use Survey required 5/9 of the sample be allocated to weekdays and 2/9 of the sample allocated to each of Saturday and Sunday. Interviewers were assigned either a 'pair' workload consisting of two selected Collection Districts (CDs) with specified days in successive weeks, or a 'single' workload consisting of one selected CD. A selected CD was assigned two specific consecutive days for enumeration, and thus all selected persons in the same CD filled in TUS diaries for the same two days. There were seven different types of 'pair' workloads, designed and randomly allocated to 'paired' CDs in such a way as to ensure equal allocation of the seven days. 'Single' workloads were then used to increase the proportion of Saturdays and Sundays, as those CDs were always enumerated on the middle weekend of the collection period. 

To enable an acceptable level of accuracy and reliability to be achieved after allowing for sample loss (through factors such as vacant dwellings inadvertently selected in the sample, non-contacts and persons out of scope and coverage) about 6,600 dwellings were selected. This number also took into account the expected rate of non-response, as determined from the 1997 survey.

Response rates

When enumeration commenced some dwellings selected for inclusion in the sample were found to have no possibility of delivering a survey response. Collectively these are referred to as sample loss, and are composed of the following groups: 

  • dwellings which are out of scope of the survey;
  • dwellings which are under construction, demolished, or converted to non-private dwellings or non dwellings;
  • private dwellings which are vacant;
  • private dwellings that contain out of scope residents (e.g. dwellings occupied by foreign diplomats and their dependants); or
  • private dwellings that contain only visitors.

The total sample loss was 886 dwellings, 13.4% of the initially selected sample.

Households selected for inclusion in the survey can be categorised as responding or non-responding households. Responding households are either fully responding or partially responding. In the Time Use Survey, some information missing from partially responding households was imputed, as described below in 'Data processing'. 

Non-responding households include: 

  • households affected by death or illness of a household member;
  • households in which the significant person(s) in the household did not respond because they could not be contacted, had language problems or refused to participate; and
  • households in which the significant person(s) did not respond to key questions.

The following table shows the response rates:

3.1 Response rates, households
 Number of households
Fully responding3 936
Partly responding191
Not responding1 599
Sample loss886
TOTAL6 612

From the fully and partly responding households, 8,442 persons aged 15 and over were in on scope and coverage for the survey. Fully completed or partially completed questionnaires (including both interview information and diaries), were obtained for 7,672 persons, 90.9% of all persons in on scope.

Number of records on the final file

To maximise the amount of activity data captured, incomplete records with one complete day have been included in the final file. Diary days were considered complete if there were activities recorded for at least 14 hours of the day. Persons with less than 14 hours of useable information for both diary days were dropped from the file. A small number of persons were also deleted from the file when complete household details were not available. Overall 711 persons were dropped from the file because they had no useable diary information or no useable interview details. However, the details of all persons with complete interview information were included in the household and family level derivations as these were derived prior to the deletion of these persons so that correct family and household information would be included on the file. 

The number of respondents that were kept on the final file was 6,961 persons (82.5% of all persons surveyed), 3,816 families and 3,643 households contributing 13,732 diary days. Of these, 50.5%(6,941) were for the first specified day and 49.5%(6,791) were for the second specified day. 2.7% of respondents provided only a single diary day. 

The quality of data from the second diary day declined very little, as shown by one of the standard measures of data quality in time use surveys, the number of episodes per day. Day one averaged 28.9 and day two 27.0 episodes per day. 

These results confirm the usefulness of the two-day diary methodology, providing nearly twice as much data as a single day diary for little extra collection cost. 

The distribution of the days of the week and weekday/weekend days on the file is as follows: 

Distribution of days
 Day OneDay TwoTotalUnweighted %Weighted %
Sunday7242 2652 98921.814.3
Monday6827171 39910.213.1
Tuesday8476621 50911.014.2
Wednesday7928231 61511.815.2
Thursday8227701 59211.614.5
Friday7658081 57311.514.4
Saturday2 3097463 05522.214.3
Weekend3 0333 0116 04444.028.6
Weekday3 9083 7807 68856.071.4
Total6 9416 79113 732100.0100.0

Data collection

Information was obtained in the Time Use Survey partly by interview and partly by self-completion diaries. Trained ABS interviewers collected information, about the household and other members of the household, from an adult member of the selected household for all persons aged 15 years or over in the household. A diary was then left for each of these persons to record their activities over two specified days. 

Survey development

Broad user consultation was undertaken during February and March of 2004. A discussion paper was used as the key instrument in obtaining feedback and input from users for the development of the survey. Testing was also carried out to investigate respondent reaction and to ensure the effectiveness of interviewing procedures and the diary format and instructions. A dress rehearsal was conducted in Perth and surrounding areas from 23 September to 5 October 2005. This test was used to: 

  • develop improvements to field procedures used in 1997;
  • trial new questions that were added to the diary and the survey interview; and
  • trial the use of the Computer Assisted Interview (CAI) which was used for the first time for the 2006 survey.

Overall, the 2006 Time Use Survey was very similar in its design and procedures to the 1997 survey with the exception of some additional questions that were added to the survey and the use of CAI for the first time. For further detail see 'Changes from previous surveys' section.

Field procedures

Selected households were initially approached by mail, informing them of their selection for the survey and advising them that an interviewer would call to arrange a suitable time to conduct the survey interview. A brochure providing some background information about the survey, information concerning the interview process and a guarantee of confidentiality was included in the initial approach letter. 

A Computer Assisted Interview was completed with information provided by a responsible adult member of the household. The interviewer instructed the contact person on when and how to complete the diary and provided a diary for each person in on scope in the household. Instructions on completing the diary and example pages were also included in the front of the diary. In addition, each person in on scope was given a letter that provided an explanation of the diary, and a small notebook and pen to allow them to record activities throughout the day. A follow-up visit was made to collect the completed diaries after the specified diary dates. 

Interviewers

Interviewers for the survey were recruited from a pool of trained interviewers with previous experience on ABS household surveys. All phases of training emphasised understanding the survey concepts, definitions and procedures in order to ensure that a standard approach was employed by all interviewers involved in the survey. Each interviewer was provided with written instructions detailing the procedures they were required to follow. 

Interviewers were allocated a workload, that is, a number of dwellings to conduct interviews and place diaries within successive weeks of the collection fortnight. Overall, workloads were smaller than usual ABS surveys to maximise the possibility of contact and placement of diaries before the days specified for diary completion.

Data collection instruments

Computer Assisted Interview

The interview was designed to be administered using standard ABS procedures for conducting interviewer surveys with a responsible adult within the household and to obtain valid and reliable results. The interview questionnaire concentrated on demographic and socio-economic information about each household person in on scope to identify population groups. To ensure consistency of approach, interviewers were instructed to ask the interview questions exactly as worded in the questionnaire. The interview included questions relating to ethnicity, education, labour force status, income, child care, age and disability, household items and household use of Internet technology. 

Diary

The diary was designed to collect information about a respondent's activities, their nature, timing and duration. All persons in on scope aged 15 years and over for selected households were asked to complete a diary for two consecutive specified days, reporting their activities in their own words. Instructions and two completed sample pages at the beginning of the diary gave respondents an idea of the type of information and level of detail required. 

The diary was divided into two separate days, showing hours with fixed intervals of five minutes covering 24 hours from 12 am. Respondents were instructed to use the five minute markings to indicate starting times, and record an arrow to the finish time for each activity. Activities were later coded by office staff using a detailed classification. Diaries were either collected by the interviewer on a return visit or they could be mailed back to the regional offices in an envelope that was provided by interviewers. Five columns with question headings organised responses into main and simultaneous activities, for whom the activity was done, where the activity took place and who else was there. 

A paper version of the Computer Assisted Interview used in the survey, and a copy of the diaries placed with respondents can be downloaded as separate pdf files from the ABS website. See Appendix 3 for details.

Data processing

A combination of clerical and computer-based systems were used to process data obtained in the survey. It was necessary to employ a variety of methods to process and edit the data which reflected the different questionnaires used to collect data from the interview and diary components of the surveys. These processes are outlined below. 

Internal system edits were applied in the CAI questionnaire to ensure the completeness and consistency of the responses being provided. The interviewer could not proceed from one section of the interview to the next until responses had been appropriately completed. 

A number of range and consistency edits were programmed into the CAI questionnaire. Edit messages automatically appeared on the screen if the information entered was either outside the permitted range for a particular question, or contradicted information already recorded. These edit queries were resolved on the spot with respondents. 

Data from the CAI questionnaire were electronically loaded to the processing database on receipt in the ABS office in each State or Territory. Computer assisted coding was performed on responses to questions on country of birth, occupation and industry of employment to ensure completeness. Data on relationships between household members were used to delineate families within the household, and to classify households by type. An outline of the computer assisted coding that was performed is provided below. 

Language spoken and country of birth coding

Occupation and industry coding

Family relationship coding

Diary processing

Unlike the household questionnaire, the activity diaries required an intensive clerical process. 

Processing of the diaries involved sorting the reported activities into episodes, editing where necessary and recording episodes into a data entry system. An episode can contain the following elements: 

  • start and finish time;
  • primary activity;
  • secondary activity;
  • person or group for whom the activity is done;
  • location, both physical and spatial;
  • mode of transport for travel items;
  • technology/communication code where relevant;
  • who the respondent was with; and
  • the age and health details of any household people present.

A change in any of these elements created a new episode.

Coding rules

Overriding activities

Priority activities

Pervasive activities

Omitted activities

People present during the activity

Diary coding system

The diary coding system used for the 2006 Time Use Survey was similar to the diary processing system used for the 1997 survey. 

The data entry system contained a look-up list of activities and detailed category screens to make it easier for consistency in coding. There were also some interactive range and logical edits. These detected unacceptable values and ensured that certain fields were appropriately coded. This type of edit, however, could not ensure that activities were coded accurately or that identifying information was correctly entered. The main quality control measures for diary information were training of coding staff, consultation and documentation to establish consistent interpretations, and audits on coding and keying. 

A quality control system was devised to ensure that coding staff were accurately following the rules laid down in the comprehensive set of coding documentation. As the quality control system involved the duplication of diaries by different coding staff, it was easy to check that coding was consistent. 

Diary information was checked by allocating a certain proportion of a coder's diaries to be recoded by another coder. Differences between the coders' data were compared and feedback was provided to each coder on a regular basis to help improve the quality of their work and improve their familiarity with the coding rules. Only the more accurate data were used towards the final output. Initially 100% checking was carried out, diminishing as reliability was demonstrated.

Additional editing

A range of processes were applied to the diary information to check that specific conditions were correctly coded according to the coding rules; that errors had not occurred in coding; and that relationships between household and diary information were consistent. A Query Resolution System was also used to ensure that: 

  • an accurate record of decisions was kept;
  • coding of episodes was consistent;
  • the Activity Classification was updated for unusual or unknown situations; and
  • coders could continue to process diaries if they could not resolve an issue within a short time.

A range of edits was also applied to the household, individual and diary information to double check that logical sequences had been followed in the questionnaires; that specific values lay within expected ranges; and that relationships between items were consistent. Unusually high values for the continuous variables such as income, the number of hours worked and duration of activities were investigated to determine whether there had been errors in entering the data.

Imputation for missing values

Some households did not supply all the required information but supplied sufficient information to be retained in the sample. Where there was income data missing from the questionnaire this could be imputed in some cases. For the Time Use Survey missing income data was imputed in two ways. Firstly, missing data for income relating to government pensions and allowances was clerically imputed by looking at the other household information and using the guidelines from Centrelink and Department of Veterans Affairs to determine the missing value. Secondly, missing data for wages and salary income was imputed by replacing each missing value with a value reported by another person (referred to as a donor). 

Donor records were selected by finding fully responding persons with information on various characteristics, such as state, sex, age, labour force status and income, that matched the person with missing information. As far as possible, the imputed information is an appropriate proxy for the information that is missing. Donors were randomly chosen from the pool of individual records with complete information for the questions where the missing information occurred.

Creation of a hierarchical file

There were two data files which fed into the combined dataset for TUS, one that included information from the interview and one that contained diary information. These two files were merged into a combined dataset featuring six levels: 

  • the top level contains household information, including items such as occupation of the household reference person and whether the household has Internet access;
  • the second level contains family information (a household can contain multiple families), including items such as the age of the youngest child in the family and whether child care was used;
  • the third level contains person information for each person aged 15 years or over in the household/family such as marital status and types of income;
  • the fourth level is the "diary day" level, where the diary information begins. Most people have two diary days of information. This level contains items about the quarter of diary collection and the day of the week that the diary day pertains to;
  • the fifth level contains the information about diary "episodes" including when the episode started and finished and where the episode took place; and
  • the sixth level then breaks each episode into primary and secondary activities (where the person stated they were doing two things at the same time, such as cooking dinner and watching TV). It contains the information about what was happening during each episode (sleeping, eating, working and so on).

Weighting, benchmarking and estimation

Weighting

Benchmarking

Estimation

Data quality

Reliability of estimates

Although care has been taken to ensure that the results of this survey are as accurate as possible, there are certain factors which affect the reliability of the results to some extent, and for which no adequate adjustments can be made. These are known as sampling error and non-sampling error. These factors, which are discussed below, should be kept in mind when interpreting the results of the survey. 

Comparisons between estimates from surveys conducted in different periods, for example, comparison of 2006 TUS estimates with 1997 TUS estimates, are also subject to the impact of any changes made to the way the survey was conducted (see Chapter 5 'Changes from previous surveys'). 

Sampling error

Sampling error is a measure of the variability that occurs by chance because a sample, rather than the entire population, is surveyed. Since the estimates in the Time Use Survey publication are based on information obtained from occupants of a sample of dwellings they are subject to sampling variability. That is, they may differ from the figures that would have been produced if all dwellings had been included in the survey. One measure of sampling variability is the standard error (SE). There are about two chances in three that a sample estimate will differ by less than one SE from the figure that would have been obtained if all dwellings, in the population described, had been included in the survey, and about nineteen chances in twenty that the difference will be less than two SEs. 

Another measure of the likely difference is the relative standard error (RSE), which is obtained by expressing the SE as a percentage of the estimate: 

\(R S E \%=\frac{S E}{e s t i m a t e} \times 100\)

The RSE is a useful measure in that it provides an immediate indication of the percentage errors likely to have occurred due to sampling, and thus avoids the need to refer also to the size of the estimate.

RSEs for estimates from the 2006 Time Use Survey are published for the first time in 'direct' form. Previously, a statistical model was produced that related the size of estimates with their corresponding RSEs, and this information was displayed in a 'SE table'. For 2006, RSEs for TUS estimates have been calculated for each estimate and published individually. The group jackknife method of variance estimation is used for this process, which involved the calculation of 60 'replicate' estimates based on 60 different subsamples of the original sample. The variability of estimates obtained from these subsamples is used to estimate the sample variability surrounding the main estimate. 

In the tables in this publication, only estimates (numbers, percentages, participation rates and means) with RSEs less than 25% are considered sufficiently reliable for most purposes. However, estimates with large RSEs (between 25% and 50%) have been included and are marked with a cell comment to indicate they have a relative standard error of 25% to 50% and should be used with caution. Estimates with RSEs of 50% or more are marked with a cell comment to indicate that they are subject to sampling variability too high for most practical purposes. 

Standard errors of proportions and percentages

Proportions and percentages formed from the ratio of two estimates are also subject to sampling error. The size of the error depends on the accuracy of both the numerator and the denominator. The RSE of a proportion or percentage can be approximated using the formula: 

\(R S E\left(\frac{x}{y}\right)=\sqrt{[R S E(x)]^{2}-[R S E(y)]^{2}}\)

This formula is only valid when \(x\) is a subset of \(y\) .

The SE of an estimated percentage or rate, computed by using sample data for both numerator and denominator, depends on the size of both numerator and denominator. However, the formula above shows that the RSE of the estimated percentage or rate will generally be lower than the RSE of the estimate of the numerator. 

Standard errors of differences

The difference between two survey estimates (of numbers or percentages) is itself an estimate and is therefore subject to sampling variability. The SE of the difference between two survey estimates depends on their SEs and the relationship (correlation) between them. An approximate SE of the difference between two estimates can be calculated using the formula: 

\(S E(x-y)=\sqrt{[S E(x)]^{2}+[S E(y)]^{2}}\)

While this formula will only be exact for differences between separate and uncorrelated (unrelated) characteristics or sub-populations, it is expected to provide a good approximation for all of the differences likely to be of interest in this publication.

Testing for statistically significant differences

For comparing estimates between surveys or between populations within a survey it is useful to determine whether apparent differences are 'real' differences between the corresponding population characteristics or simply the product of differences between the survey samples. One way to examine this is to determine whether the difference between the estimates is statistically significant. This is done by calculating the standard error of the difference between two estimates (x and y) and using that to calculate the test statistic using the formula: 

\(\frac{|x-y|}{S E(x-y)}\)

If the value of the test statistic is greater than 1.96 then we may say that we are 95% certain that there is a statistically significant difference between the two populations with respect to that characteristic. Otherwise, it cannot be stated with confidence that there is a real difference between the populations.

Non-sampling error

Lack of precision due to sampling variability should not be confused with inaccuracies that may occur for other reasons such as errors in response and recording. Inaccuracies of this type are referred to as non-sampling error. This type of error is not specific to sample surveys and can occur in a census enumeration. The major sources are: 

  • errors related to scope and coverage;
  • response errors such as incorrect interpretations or wording of questions,
  • interviewer bias;
  • non-response bias; and
  • processing errors.

Errors related to scope and coverage

Some dwellings may have been inadvertently included or excluded because, for example, the distinctions between private and special dwellings were unclear. Some persons may have been wrongly included or excluded because of difficulties applying the coverage rules concerning, for instance, household visitors, or scope rules concerning persons excluded from the survey. Particular attention was paid to question design and interviewer training to ensure such cases were kept to a minimum.

Response errors

Response errors may have arisen from three main sources: 

  • deficiencies in questionnaire design and methodology;
  • deficiencies in interviewing technique; and
  • inaccurate reporting by respondents.

Errors may be caused by ambiguous or misleading questions, or, in the case of diaries, ambiguous column headings or example pages, inadequate or inconsistent definitions of terminology used, or by poor questionnaire sequence guides, causing some questions to be missed. Thorough testing occurred before the questionnaire and diary format were finalised to overcome problems in questionnaire and diary content, design and layout.

Lack of uniformity in interviewing also results in non-sampling error. Thorough training programs, a standard Interviewer's Manual, the use of experienced interviewers and checking of interviewers' work were methods employed to achieve and maintain uniform interviewing practices and a high level of accuracy in recording answers on the survey questionnaire. 

A respondent's perception of the personal characteristics of the interviewer can be a source of error. The age, sex, appearance or manner of the interviewer may influence the answers obtained. 

In addition to the response errors described above, inaccurate reporting by respondents may occur due to misunderstanding of the question, inability to recall the required information and deliberately incorrect answering to protect personal privacy. 

Non-response bias

One of the main sources of non-sampling error is non-response when persons resident in households selected in the survey cannot be contacted, or if they are contacted are unable or unwilling to participate. Non-response can affect the reliability of results and can introduce bias. The magnitude of any bias depends upon the level of non-response and the extent of the difference between the characteristics of those people who responded to the survey and those who did not. For the 2006 TUS, some of the non-response resulted from logistical difficulty in aligning interview times with allocated diary days rather than the unwillingness of selected household members to participate in the survey. 

As it was not possible to quantify accurately the nature and extent of the differences between respondents and non-respondents in the survey, every effort was made to reduce the level of non-response. The estimation procedures used make some adjustment for non-response. 

Processing errors

Processing errors may occur at any stage between initial collection of the data and final compilation of statistics. There are four stages where error may occur: 

  • coding, where errors may have occurred during the coding of various items by office processors;
  • data transfer, where errors may have occurred during the transfer of data from the questionnaires to the data file;
  • editing, where computer editing programs may have failed to detect errors which reasonably could have been corrected; and
  • manipulation of data, where inappropriate edit checks, inaccurate weights in the estimation procedure and incorrect derivation of new items from raw survey data can also introduce errors into the results.

Steps to minimise errors

A number of steps were taken to minimise errors at various stages of processing. These included:

  • thorough training of staff;
  • providing detailed coding instructions and regular checking of work performed;
  • computer edits designed to detect reporting or recording errors;
  • validation of the data file using tabulations to check the distribution of persons for different characteristics; and
  • investigation of unusual values on the data file.

Data quality issues specific to time use

The stages where error may be introduced, as listed above, also apply to diary data. There are, in addition, a number of other data quality issues which could affect the data from self-completion diaries, from specific coding problems to broader conceptual questions. 

The use of a diary in which people record their activities by time of day for two specified days was a choice based on considerable research, testing and evaluation in Canada, Europe and Australia. The ABS has tested and evaluated a range of methodologies through the Time Use Pilot Survey conducted in Sydney in 1987 and smaller pilot tests for the 1992 and 1997 TUSs. 

There are two possible choices with self-completion diaries: to ask people to select from a pre-coded list of activities; or to ask them to describe their activities in their own words. 

The use of a pre-coded list limits the number of categories of activities, to avoid confusing respondents. This can provide useful information when the survey is intended for a particular purpose, where tables are usually presented with a relatively small number of categories. In addition, processing diaries completed from a pre-coded list is a relatively simple exercise. 

Inviting people to record activities in their own words has the advantage that: 

  • people are able to put their own priority on the activities they select to report;
  • the more detailed activities people report can provide feedback on emerging trends, e.g. recycling activities, which may be useful for later classifications; and
  • there is less of a 'leading question' effect, prompting a socially desirable answer.

The open ended survey was chosen by the ABS because the greater detail collected and stored will meet the needs of a wider variety of users and allow them to aggregate items according to their own purposes.

Reporting variability

Number of episodes

Classification issues

Limitations on data items

Education

Highest qualification may in some cases include qualifications not accredited in Australia, including overseas qualifications. 

Employment

A short labour force module was used. Users should refer to Labour Force, Australia (Cat. no. 6202.0) for more detailed information on employment. The population included in the question on 'reasons not looking for work' is very small. It is not intended to provide data on discouraged job seekers; the main reason for asking this question is to identify those constrained to stay out of employment by family responsibilities, and those who prefer to stay home while children are young. 

As part of the labour force module in the interview part of the survey, respondents are asked to record their main activity. This question assesses respondents self perception of their main activity and does not use any strict definitions. Therefore, a respondent could say that their main activity is 'being ill or disabled'. However, in the disability module they may not indicate that they have any conditions lasting for six months or more. Furthermore, this question asks about the respondents 'current' main activity and therefore they may have indicated that they did not work last week in the Labour Force module but still indicate that their current main activity is working. 

Income

As the information was collected from a responsible adult in the household about other members, a limited subset of the more detailed income questions was used. 

Disability and assistance

The module of data items about disability is not designed to provide counts of people by disability status. The ABS conducts a regular large scale survey on this topic, the Survey of Disability, Ageing and Carers, the most recent in 2003 (Cat. no. 4430.0). The purpose of this module is to allow the analyst to study the effect of disability on the person's and the household's use of time. 

The data item relating to assistance given to children who have a condition lasting six months or more should be used with care for children under five years because of the difficulty of separating the assistance needed or given on account of the child's age from that given because of disability. Again, the purpose of this item is to show the impact of a child's disability or long-term health condition on the activity patterns of parents and other household members. 

Care should also be used with the data item relating to why the household member gives assistance to a person outside of the household. The assessment that the person receiving assistance has a disability is made by the person responding and may not be medically correct or correspond with disability as collected by the disability module on the survey form. 

Child care

The data items about child care were not designed to be used for information about children using child care or the adequacy of child care provision. Fewer questions were asked than in the ABS Child Care Survey, which is a regular large scale survey conducted by the ABS on this topic, see Child Care Australia (Cat. no. 4402.0). These items were included to assist in interpreting the parent's patterns of time use, as the use of child care frees a parent's time for other activities. 

For whom

In 2006 an additional category was added for the coding of the 'for whom' column of the diary to include activities that were done for family members within the household who were ill or with a disability. This can be used to some extent to look at the time spent caring for family members who are ill or have a disability. Activities were coded as being for someone who was ill or with a disability based on information within the diary and the interview. If household members indicated in the disability module of the interview that they had a condition that was likely to last for six months or more then activities that were done for these household members would be coded as being for a family member who was ill or with a disability. It cannot be determined whether the activity was being done for a respondent because of their disability or whether it was normal exchange between household members that is unrelated to the disability. For example, someone could be cooking dinner for themselves and their family. If one person in their family happens to have a condition lasting for six months or more (for example asthma) this activity would have been coded as being for 'family who are well' as well as for 'family with a disability'. Even though the family member has asthma, they most likely would still have been able to cook their own dinner. Due to this there are limitations on the way this item can be used to provide information about the care given to household members with a disability. However, the category provides the potential to capture conversations for emotional support, or TV watching as a form of supervision. The most useful way to study the activities of carers is to look at the overall balance between their own personal care and other activities. 

Secondary activities

For primary activities, information is captured about who the activity was done for as well as the type of communication and technology used during the activity (if any). For secondary activities this supporting information is not captured. Therefore, assumptions cannot be made about the purpose of these activities. When looking at use of computers a distinction between the different types of computer use could only be made for primary activities. For example, for primary activities there is a distinction between computers used for communication and for computer use that does not involve communication. There is also a distinction between computer use without the Internet and computer use that involves the Internet. None of these distinctions can be made if the computer use was occurring as a secondary activity.

Changes from previous surveys

Overview

The 2006 survey was designed to be as comparable as possible with the 1997 survey. However, the 2006 Time Use Survey did introduce some changes. The changes were largely designed to improve survey quality but may impact on the comparability between the 2006 estimates and earlier data. Notwithstanding these changes, a high level of comparability with 1997 survey results has been achieved. This chapter outlines the main changes.

Changes to the 2006 survey instruments

The 2006 survey introduced a number of data items that had not been collected in previous surveys. The main additions were: 

  • questions in the diary about:
    • self assessed health status;
    • trust of people;
    • satisfaction with the time spent on the diary days;
    • satisfaction with time spent alone or with other people;
    • attitude towards gardening; and
    • volunteer work.
  • questions in the questionnaire about:
    • household use of computers and the Internet;
    • proficiency in spoken English;
    • indicators of casual employment;
    • working from home as an arrangement with an employer, and technology used to work from home;
    • whether the dwelling is managed by a body corporate; and
    • whether respondents are grandparents and whether they provide care for their grandchildren.

Changes to the classifications and coding lists

Activities

The same nature and purpose activity classifications were used for the 2006 survey as was used for 1997 (with the exception of some minor revisions). There were, however, significant changes between the 1992 Activity Classification and the 1997 and 2006 classifications. A concordance data item has been built applying the 'for whom' and 'communication/technology' data items to the 1997 and 2006 activity classifications to achieve a reasonable level of comparability with the 1992 classification. 

Communication and Technology

Communication in 1997 and 2006 was coded in a different way from 1992. In 1997 and 2006, the communication/technology supplementary classification allows the coding of communication episodes to the activity being discussed. For example, if the respondent recorded that they phoned their daughter to ask her to bring in the washing, this would be coded to 'bringing in and hanging out the washing' with a communication/technology code of 'phone'. To allow for direct comparisons with 1992, the 'purpose of the activity' data item was amended to move these episodes of communication to correspond with the 1992 communication codes. 

Communication in 2006 was coded in a different way from 1997. There were additional categories used to capture different forms of communication and technology that may have been more commonly used in 2006 including mobile phones and the Internet. The coding list was also split into two groups so that activities involving communication were coded differently to activities that utilised technology but were not related to communication. For example, there was a different code used to distinguish communication via the Internet (such as email) than for an activity involving the use of the Internet (such as doing research) that does not involve communication. This allowed all technology use to be captured for the primary activities instead of only collecting information on the technology used for communication. Communication and technology codes were only coded for primary activities. 

For whom

The 'for whom' column records an important aspect of each activity: its purpose. This column identifies the activities people have done for themselves and those done for other people or organisations. The 2006 coding list for this variable was slightly different to the 1997 coding list. In 2006, activities that were done for family members within the household were coded differently depending on whether the household members were sick, frail or with a disability or whether they were well. In 1997 there was not a distinction between whether the activity was done for a family member in the household who was well or whether it was done for a family member who was sick, frail or with a disability. This information was only collected and coded for the primary activity. Secondary activities do not have any information associated with them about the purpose of the activity. 

Location

Location is coded for both the physical location (e.g. own house or workplace) and the spatial location (e.g. indoors or outdoors). For 2006, an additional category was added for physical location to distinguish time spent at beaches, rivers, or lakes from other country settings. Furthermore, an additional category was added to include a spatial location of waiting in cars separate from waiting indoors or outdoors.

Coding rule changes

There were some differences in the coding rules that were used to code the diaries that may have some impact on the comparisons made between the 2006 data and the 1997 data. Some coding rules were changed in 2006 to improve the consistency of the data, the major changes are outlined below. 

Child care

Talking/playing/reading to children is an activity category within the Activity Classification. There was a slight change in the way that talking activities were coded to this category. In 1997, when a respondent reported 'talking to family' and there were children under 15 as well as adults present this was usually coded to 'talking for recreation and leisure' and no time was allocated to 'talking to children' unless it was clear that the child was involved. However, in 2006 episodes of 'talking to family' were often split into two episodes to include the time for talking to the adults and talking to the children when there were children under the age of 15 present. The 'talking to children' was coded as the primary activity and the 'talking to adults' was coded as the secondary activity. This was because child care activities were seen as priority activities. In 1997, 'talking to children' was not always seen as a priority activity and was therefore not always recorded as the primary activity. This may have some impact on the comparisons between the time spent on this activity in 1997 and in 2006. 

Communication

Communication codes were used more consistently throughout the coding process in 2006. For certain activities such as: talking to adults or children; playing or reading to children; socialising; purchasing in a commercial area; and computer use, communication/technology codes were used unless there was specific information in the diary to the contrary. No communication/technology codes were used for employment activities because the majority of people did not provide detail in the diary about their employment activities and therefore no assumptions could be made about the type of communication or technology that was being used during these episodes. 

For whom

The 'for whom' column of the diary was also coded more logically in 2006. All personal care activities such as sleeping, eating or personal hygiene were coded as being 'for self'. This was because logically a person cannot be doing these activities for someone else. In 1997 this was not always the case. Some personal care activities such as sleeping and eating were coded as being for 'family' or other people. In 2006, although some respondents may have recorded that they were doing these activities for people other than themselves this was edited during the processing of the diaries so that a consistent logical rule was applied for the personal care activities of sleeping, eating and personal hygiene. 

Furthermore, all employment related activities for the main job or other jobs (excluding work breaks) were coded as being 'for work'. Many people recorded that the activity was done for themselves or their family. Once again these were changed during the processing of the diaries to apply a consistent rule across the work episodes. In 1997 this was not the case. 

Purchasing

There were some differences in the way that purchasing activities were coded in 2006 compared with 1997. In 1997, activities such as eating at a cafe/restaurant, or having drinks at a pub or nightclub were coded entirely to either eating or social/alcoholic drinking. However, in 2006 the first five minutes of the episodes at the eating or drinking locale were coded to purchasing consumer goods and the remainder was coded to either eating or drinking. 

Travel

Furthermore, travel to and from all eating and drinking locales was coded to travel associated with purchasing. In 1997, travel to nightclubs or pubs was coded to travel associated with recreation and leisure. This may contribute to an increase in the amount of time spent on travel associated with purchasing and a decrease in the amount of time spent on recreation and leisure. 

In 1997, travel to and from work was mostly coded as travel associated with employment related activities unless the activity was one of the priority activities such as domestic, child care or caring for adults/doing volunteer work. However, in 2006 all travel was coded to the purpose of the activity. For example, if someone dropped their child at child care on the way to work, the travel from home to the child care centre was coded to travel associated with child care and the travel from the child care centre to work was coded as travel associated with employment. However, in 1997 this would all have been coded to travel associated with employment activities. This may have an impact on the comparability of the estimates for travel activities.

Changes to processing procedures

There are three different versions of the Activity Classification data, the nature of the activity, the purpose of the activity and the activity as concorded to the 1992 classification. The purpose and concordance versions of the data were derived from the nature of the activity using other supporting information such as who the activity was done for and whether communication was associated with the activity. There are differences between the data for each of these versions of the activity data. In 2006, there was a change made to the purpose derivation to include additional activity codes in the derivation of care for adults. This has resulted in an increase in the amount of time spent on care of adults according to the purpose derivation. This change is not reflected in data from the concordance derivation as this was kept in line with how the data was derived in 1997.

Survey output and dissemination

Overview

The range of data available from the 2006 Time Use Survey in both published and unpublished form is described below. More detailed information can also be obtained by telephoning the Family and Community Statistics section on (02) 6252 7880.

Publications and catalogued releases

Releases for 2006 survey

Special data services

Releases from previous surveys

Related publications

Using the CURF

Introduction

This chapter provides information about the basic and expanded Confidentialised Unit Record Files (CURFs) from the 2006 Time Use Survey (TUS). Two microdata files are available from this survey: 

  • a basic TUS CURF available on CD-ROM or through the Remote Access Data Laboratory (RADL); and
  • an expanded TUS CURF accessible only through the RADL.

The expanded CURF contains more detailed data for some variables than the basic CURF, as well as some additional variables.

The RADL is a secure on-line data query service, that clients can access via the ABS website. Because the CURFs are kept within the ABS environment, the ABS is able to release more detailed data via the RADL than can be made available on CD-ROM. Further information about this facility and obtaining access to the CURFs is available on the ABS Website (see Services We Provide/CURF Microdata/Accessing CURF Microdata). 

In addition to this chapter, TUS CURF users should also use the information in other chapters of this User Guide.

About the microdata

The 2006 TUS CURFs contain unit records relating to almost all survey respondents. 

The data are released under the Census and Statistics Act 1905, which has provision for the release of data in the form of unit records where the information is not likely to enable the identification of a particular person or organisation. Accordingly, there are no names or addresses of survey respondents on the CURFs and other steps have been taken to protect the confidentiality of respondents. These include: 

  • removing persons from large households on the file to reduce them to a maximum household size of 8 persons on the expanded CURF and 6 persons on the basic CURF;
  • reducing the level of detail for many data items (e.g. geography, industry and occupation items);
  • perturbing income items or presenting deciles only;
  • ranging, collapsing or top-coding some variables (e.g. age of person); and
  • changing demographic information for a small number of records.

As a consequence, aggregated data obtained from the CURFs are slightly different to that published in How Australians Use Their Time (Cat. no. 4153.0). See the section 'Reconciliation of the data' in this chapter for more information.

Steps taken to confidentialise the datasets made available on the CURFs are undertaken in such a way as to ensure the integrity of the datasets and optimise the content, while maintaining the confidentiality of respondents. Intending purchasers should ensure that the data they require, at the level of detail they require, are available on the CURFs. Data collected in the survey but not contained on the CURFs may be available in tabulated form on request. For a complete list of the data items and categories on the basic and expanded CURFs see Appendix 2.

Data items

The detailed data item list and categories for the TUS CURFs is contained in Appendix 2. The data item list contains an index based on subject and an index based on field name. The data items included on the CURFs, and the categories within the data items, differ between the basic and expanded CURFs. The expanded CURF contains more variables than the basic CURF as well as more detailed data for selected variables. Table 8.1 shows the differences between the 2006 basic and expanded TUS CURFs. Most of the differences result from the difference in the level of detail and the difference in the maximum household size permitted on the basic and expanded CURFs. On the basic CURF, households with 7 or more members have been reduced to a maximum of 6, while on the expanded CURF households with 9 or more members have been reduced to a maximum of 8.

Table 8.1 Comparison of data between 2006 basic and expanded TUS CURFs

Notes on specific data items

Geographic items

To enable CURF users greater flexibility in their analyses, the ABS has included one Socio-economic Index for Area (SEIFA) and several sub-state geography items on the expanded 2006 CURF. Conditions are placed on the use of these items. Tables showing multiple data items, cross tabulated by more than one sub-state geography at a time, are not permitted due to the detailed information about small geographic regions that could be presented. However, simple cross-tabulations of population counts by sub-state geographic data items may be useful for clients in order to determine which geography item to include in their primary analysis, and such output is permitted. 

Income items

Continuous income items (expanded CURF)

The person level records contain information on income by source (continuous) and total weekly income (continuous and in deciles). The household level records contain information at a broader level, derived from the income of all persons aged 15 years and over in the household (household gross weekly income and equivalised gross income are continuous and in deciles). Users should use the household income variables, as persons included in family and household level variables may not have person level records on the CURF files (refer to section 'Persons 15 years and over' in this chapter). As a result, users deriving family or household level income measures from the person level file will not derive the correct amounts. 

When analysing income at person level, it is necessary to exclude the reserved values of 99,999,998 for 'Not known' and 99,999,999 for 'Not stated' for income items for specific sources. When analysing income totals at person and household level, it is necessary to exclude the reserved value of 99,999,998 for 'Not known or not stated'. Note that if more than one contributing income item at the person level has a value of 'Not known' or 'Not stated', then 'Total weekly cash income' and derived deciles are set to 'Not known or not stated', as it was not possible to derive an accurate total. Similarly, if more than one contributing person record in a household has a value of 'Not known or not stated', then household income (both equivalised and gross), and derived income deciles are set to 'Not known or not stated'. 

The person level income items are: 

  • weekly income from unincorporated business (OINCO05A);
  • weekly income from wages and salary (OINCO11A);
  • weekly income from government cash pensions (OINCO11B);
  • weekly income from rental income (OINCO08);
  • weekly income from dividends and interest (DIVINTWK);
  • weekly income from other cash sources (OINCO10); and
  • total weekly cash income (OINCO12).

The relevant household level income items are:

  • household equivalised gross weekly income (EQUIVINC); and
  • household gross weekly income (OINCO16).
Income deciles (expanded and basic CURFs)

When analysing income deciles at person or household level, it is necessary to exclude the reserved value of 00 for 'Not known or not stated'. Refer to the section above 'Continuous income items (expanded CURF)' for more details on how the 'Not known or not stated' values are derived from individual income components.

The person and household level income items in deciles are: 

  • Total weekly cash income deciles (PERSDEC);
  • Household equivalised gross weekly income (EQUIVDEC); and
  • Household gross weekly income deciles (HHDEC) (expanded CURF only).

Multiple response items

There are two data items at the person level on the 2006 TUS CURFs that have multiple responses. In these instances respondents were able to select one or more response categories, and the output data items are multi-response in nature. This section describes these items and provides some information on how to use them.

On the basic and expanded CURFs, the data items are: 

  • relationship of carer to recipient living in household (RELCRHA-RELCRHC); and
  • relationship of primary carer to recipient living in household (RLPCRHA-RLPCRHC )

The item 'Relationship of carer to recipient living in household' captures multiple responses where a person provides care to more than one person with a disability in their household. The first response is captured in the first, or 'A', position (RELCRHA), and additional responses are in the second and then third, or 'B' and 'C', positions (RELCRHB, RELCRHC). If a person does not care for anyone then they will have a value of 99 'not applicable' in the first position (RELCRHA). The 'Null response' (value of 0) is a default code and should not be used in data analysis. The item 'Relationship of primary carer to recipient living in household' is set up in the same way.

Imputation flags

Imputation was undertaken for income from wages and salary income only, and an imputation flag (IMPFLAGP) exists at the person level to indicate whether an individual record had income from wages and salary imputed. A value of 1 'partially imputed' indicates that income from wages and salary for that record was imputed. On the expanded CURF this flag can be used to determine if 'Weekly income from wages and salary', and also if 'Total weekly cash income' includes an imputed wages and salary component. On the expanded and basic CURFs, the flag can also be used to indicate if 'Total weekly cash income deciles' includes an imputed wages and salary component. An imputation flag also exists at the household level (IMPFLAGH) to indicate whether a household contains one or more person records which had a wages and salary item imputed, and this can be used in a similar manner.

Deciles

Table 8.2 Total weekly cash income decile boundaries (PERSDEC) for Expanded CURF

Table 8.3 Household equivalised gross weekly income decile boundaries (EQUIVDEC) for Expanded CURF

Table 8.4 Household gross weekly income decile boundaries (HHDEC) for Expanded CURF

Decile boundaries are similar for the Basic CURF but have not been published.

File structure and use

Nature of the levels

Each of the CURFs contain the following record levels:

  • Household level - contains information about State or Territory and area (Capital City/Balance of State) of residence, type of dwelling, tenure type, landlord type, household type and composition, household income and number of income earners, disability, household items (such as televisions and vehicles), household use of Internet technology, and some information relating to the household reference person;
  • Family level - contains information about family composition, child care, disability, and some information relating to the family reference person;
  • Person level - contains information about each selected person, the family and household to which they belong, such as: sex, age, marital status, relationship in household, country of birth, year of arrival in Australia, family type, income unit type, labour force details, occupation and industry, education status, education qualifications and educational institution attending, income by detailed source of income, child care, carer and disability information, use of Internet technology, time use. Person records only exist for persons aged 15 and over;
  • Day level - contains information about the nature of the day, the quarter of collection and the day of the week the diary information related to;
  • Episode level - contains information about each segment of time throughout the day, such as: who was present, length of episode, start and stop time of episode, who activity was done for, health information about person(s) present, spatial and physical location of episode, mode of transport used, and type of communication/information technology used during the episode as recorded in the TUS diary; and
  • Activity level - contains information about what was happening during each episode such as the nature and purpose of activity as recorded in the TUS diary, and an activity concordance to 1992 Time Use Survey. It also provides information about whether the activity was a primary or secondary activity (where the person stated that they were doing two things at the same time, such as cooking dinner and watching television).

Table 8.5 shows the number of records on each level.

Table 8.5 Record counts
 TUS Expanded CURFTUS Basic CURF
Household level3 6433 626
Family level3 8153 793
Person level6 9606 902
Day level13 73013 617
Episode level384 484381 355
Activity level520 441516 219

Identifiers

There are seven identifiers on the expanded and basic CURFs.

Every household has a unique random identifier (ABSHID). This identifier appears on the household level, and is repeated on the family, person, day, episode and activity levels for each record relating to that household. 

Each family within a household is numbered sequentially commencing at 1. Non-family members, single person households and persons in group households have a sequential "family number" commencing at 50. Family number (ABSFID) appears on the family level and is repeated on the person, day, episode and activity levels for each record relating to that family. The combination of household and family number uniquely identifies a family. 

A family has one or more income units and each income unit within the family is numbered sequentially. Income unit number (ABSIID) appears on the person, day, episode and activity levels. The combination of household, family, and income unit number uniquely identifies an income unit. 

An income unit has one or more persons and each person within the income unit is numbered sequentially. Person number (ABSPID) appears on the person level and is repeated on the day, episode and activity levels. The combination of household, family, income unit and person number uniquely identifies a person. 

A person has records for up to two days and each day within a person record is numbered sequentially. Day number (ABSDID) appears on the day level and is repeated on the episode and activity levels. The combination of household, family, income unit, person and day number uniquely identifies a day. 

A day has one or more episode records and each episode within a day record is numbered sequentially. Episode number (ABSEID) appears on the episode level and is repeated on the activity level. The combination of household, family, income unit, person, day and episode number uniquely identifies an episode. 

An episode has one or more activities and each activity within an episode record is numbered sequentially. Activity number (ABSAID) appears on the activity level. The combination of household, family, income unit, person, day, episode and activity number uniquely identifies an activity. 

At higher levels, level identifiers for lower levels are set to zero. 

Children aged under 15 years

Children aged under 15 years do not have their own person level record on the file. Information on the number and ages of such children was collected and is included on the household and family level files. 

Persons aged 15 years and over

In some cases persons aged 15 years and over included in family and household level variables may not have individual person, day, episode or activity records on the CURF files. These individuals either had no useable diary information or interview details and were dropped based on the rules detailed in the section 'Number of records on the final file' (see 'Survey Methodology' section). 

Use of weights

The CURFs contain records which can be adjusted (weighted) to infer results for the total in-scope population in Australia. As the survey was conducted on a sample of private dwellings in Australia, it is important to take account of the method of sample selection when deriving estimates from the CURFs. This is particularly important as a person's chance of selection in the survey varied depending on the State or Territory in which the person lived, as well as the days of the week they were asked to respond for. If the chance of selection is not accounted for, by use of appropriate weights, the results will be biased. For information about the derivation of the weights for the 2006 TUS see 'Survey Methodology' section in this User Guide. 

Each household, family, person, day, episode and activity record contains a weight. This weight indicates how many population units are represented by the sample unit. Weights for each family record are the same as the weights for the household record. Likewise, weights for the episode and activity records are the same as the weights for the person-day record. Care needs to be taken to ensure the appropriate weight is selected when estimating for the desired Australian population (see Appendix 2 for weights at each level). 

In addition, replicate weights have been included on the CURFs which can be used to calculate sampling error. Sampling error arises because the estimates are based on a sample of units and so will differ from estimates that would have been produced if all units in the population had been included in the survey. Each record on the CURFs contain 60 'replicate weights' in addition to the associated 'main weight'. Information on the use of these replicate weights is provided in the section 'Reliability of the estimates' below. 

Weights are calibrated against population benchmarks to ensure that estimates conform to an independently estimated distribution of the population by certain characteristics, rather than to the distributions within the sample itself. Separate benchmarks are used for each enumeration period, and include households and persons residing in occupied private dwellings only. The benchmarks sum to an averaged "yearly" private dwelling population, and therefore do not, and are not intended to, match estimates of the total Australian resident population published by the ABS. For information about the benchmarks used in the calibration of the final weights please see 'Survey Methodology' section in this User Guide. 

If estimates of population sub-groups are to be derived from the CURFs, it is essential that they are calculated using the appropriate weights of records in each category and not just by counting the number of records in each category. If weights for each category were to be ignored when analysing the data to draw inferences about the population of interest, then no account would be taken of the chance of selection or of different response rates across population groups. 

It should be noted that as a result of some of the changes made to protect confidentiality on the CURFs, estimates of benchmarked items produced from the CURFs may not equal the benchmarked values. Further information about this difference in estimates is presented in the section 'About the Microdata' at the start of this chapter. 

Reconciliation of the data

It is not possible to reconcile exactly the data produced from the CURFs with published data. This is a result of the steps taken to preserve confidentiality. These steps are outlined in the section 'About the Microdata' at the start of this chapter.

Using the episode and activity datasets

Introduction

The TUS 2006 CURFs comprise six datasets at different levels. One of these datasets relates to 'Episode' level data (TUS06EE for the expanded CURF or TUS06BE for the basic CURF), and another to 'Activity' level data (TUS06EA for the expanded CURF or TUS06BA for the basic CURF). Explanation of the nature and use of the data contained within these datasets is below. 

An "episode" relates to a segment of time, whereas an "activity" describes what was being done during that segment of time. 

The episode level contains information about episode start time (EPSTART) and episode stop time (EPSTOP), expressed as the number of minutes from midnight (for example, 10 am is 600), episode length (EPLENGTH), which is the duration of the episode in minutes, determined by subtracting the episode start time from the episode stop time, as well as information about where the episode took place, who else was present (their age groups and health characteristics), and who the activity was done for. 

The activity level splits each episode into primary and secondary activities. For example, a person can be cooking dinner as their primary activity and can also be watching television which would be their secondary activity. The purpose of the activity level is to explain what was being done during each episode (e.g. sleeping, eating, working). 

The activity level datasets present the information in three ways: 

The item 'Nature of activity' (NATURACT) details the type of activity based on the current Time Use Activity Classification (refer to Appendix 1). 

The item 'Purpose of activity' (PURPMN) also uses the Time Use Activity Classification but it looks at the purpose of the activity. For example, if a person was doing gardening for their elderly mother with a disability then the nature would be "gardening" whereas the purpose would be "caring for adults". 

Activity as concorded to 1992 Time Use Activity Classification (CONCORD) details the activity based on the 1992 Time Use Survey Activity Classification. 

The episode and activity levels can be used directly for all types of tables apart from tables that require ranged duration of time. Ranged duration of time looks at the total time spent by a person over all episodes which meet a given set of criteria. For example, a person who has three meal episodes in one day lasting 30 minutes each would have a total time for all episodes of eating of 90 minutes. They should therefore be counted in a ranged time table at the 90 minute point. This will only happen if you sum the episode lengths before they are ranged, otherwise, using the above example, each episode would be ranged incorrectly at the 30 minute point. 

To sum episode length

In order to sum episode length, the process has been explained in a step by step process below. Following these steps will result in the creation of a new data item that has the total time for all episodes that meet the given set of criteria: 

  • sort the episode level dataset by day. All records for a particular day must be sorted together. This can be done by sorting on the identifiers for household, family, income unit, person, and day (ABSHID, ABSFID, ABSIID, ABSPID, and ABSDID); and
  • for each record in the episode level dataset do as follows:
    • if the record is the first for a given day then set the new total field to zero. The "new total field" is the new item being created which will eventually have the total time summed for all episodes of interest;
    • if the record meets the criteria for the total then add the duration of the episode to the total. The summing criteria refer to fields on the episode level. For example, you may be interested in episodes that occur in one type of location and/or where the activity was done for a particular type of person. The duration is in minutes and is stored in the item EPLENGTH; and
    • if the record is the last for a given day then output a new episode summary record. The "new total field" item should now have summed together all the EPLENGTH values that have met the criteria for that particular day and the last record can be output with the total value.

The summing criteria can also be applied to the activity level but require the episode level fields to be copied down to the activity level before summing.

Copying data from episode to activity level

In order to copy data from the episode level to the activity level: 

  • sort the episode and activity level datasets by episode. All records for a particular episode must be sorted together. This can be done by sorting on the identifiers for household, family, income unit, person, day, and episode (ABSHID, ABSFID, ABSIID, ABSPID, ABSDID, and ABSEID);
  • match the records in the episode and activity level datasets by episode. The records must be matched using the identifiers for household, family, income unit, person, day, and episode (ABSHID, ABSFID, ABSIID, ABSPID, ABSDID, and ABSEID); and
  • add the episode level fields to the corresponding activity level records.

These steps will result in a new dataset containing all the episode level information (e.g. episode length) attached to the primary and secondary activity level data (e.g. nature of the activity). This activity level dataset can now be sorted and a "new total field" item created as per the "To sum episode length" steps detailed above. Note, however, that if you choose to include both primary and secondary activities, the algorithm must be modified to avoid double counting.

This step: 

  • if the record meets the criteria for the total then add the duration of the episode to the total

becomes:

  • if the record meets the criteria for the total and the current episode has not yet been counted then add the duration of the episode to the total

Regardless of whether you sum episode length from the episode level or from the activity level, you will end up with a day level dataset containing the "new total field". This can then be ranged and used in the same way as any other day level data item. For example, if you wanted to cross tabulate "Total duration of all episodes of eating (30 minute ranges)" by "Sex of person", you would first need to copy the latter from the person level dataset to the new day level dataset by adapting the "To copy data from the episode level to the activity level" steps above.

Reliability of estimtes

Two types of error, sampling error and non-sampling error, are possible in an estimate based on a sample survey. More information is available in 'Data Quality' section in this User Guide. 

Standard Errors (SEs) are one of the methods used to measure sampling variability. As mentioned in the section 'Use of weights' earlier in this chapter, each record on the CURFs contain 60 'replicate weights'. The SE for each estimate produced from the CURFs can be calculated using the replicate weights provided. 

SEs can be calculated using what is known as the 60 group jackknife error estimator. When calculating SEs it is important to select the replicate weights which are most appropriate for the analysis being undertaken (see Appendix 2 for replicate weights at each level). 

To obtain the SE of a weighted estimate y, calculate the same estimate using each of the 60 replicate weights. The variability between these replicate estimates (denoting y(g) for group number g) is used to measure the SE of the original weighted estimate y, using the formula: 

\(S E(y)=\sqrt{59/60 \sum \limits_{g=1}^{60}(y_{(g)}-y)^{2}}\)

where

\(g\) = the replicate group number 

\(y(_g)\) = the weighted estimate, having applied the weights for replicate group \(g\) 

\(y\) = the weighted estimate from the full sample 

Use of the 60 group jackknife method for complex estimates, such as regression parameters from a statistical model, is not straightforward and may not be appropriate. The method as described does not apply to investigations where survey weights are not used, such as unweighted statistical modelling. For more information on the 60 group jackknife method of SE estimated, see Research Paper: Weighting and Standard Error Estimation for ABS Household Surveys (Methodology Advisory Committee) (Cat. no. 1352.0.55.029). 

Information on calculating the relative standard error (RSE) of an estimate is available in 'Data Quality' section in this User Guide. 

CURF users should be aware that estimates produced from the CURFs will differ from those in the published data due to actions taken to preserve confidentiality on the TUS CURFs.

Comparison with previous CURFS

While efforts have been made to maintain comparability between CURFs where possible, changes in data collection, data item standards, analysis requirements and user preferences have resulted in changes to data items which may have an impact on data analysis and the assessment of changes over time. 

Changes impacting on all items

Changes relating to data items

Contents of the CURFs

This section provides details of the files included on each CURF, including information on the differences in files available for the basic CURF on the CD-ROM and through the RADL.

Basic TUS CURF files

Basic CURF files on CD-ROM only

Basic CURF files on RADL only

Expanded TUS CURF files

Expanded CURF test files

Conditions of release

The 2006 TUS basic and expanded CURFs are released in accordance with a Ministerial Determination (Clause 7, Statutory Rules 1983, No. 19) in pursuance of section 13 of the Census and Statistics Act 1905. As required by the Determination, the CURFs have been designed so that the information on the files is not likely to enable the identification of the particular person or organisation to which it relates. 

The Australian Statistician's approval is required for each release of the CURF. In addition, and prior to being granted access to the CURF, all organisations, and individuals within organisations, who request access to the CURF will be required to sign an undertaking to abide by the legislative restrictions on use. Organisations and individuals who seek access to the 2006 TUS basic and/or expanded CURF are required to give an undertaking which includes, among other conditions, that in using the CURF data they will: 

  • use the data only for the statistical purposes specified
  • not attempt to identify particular persons or organisations
  • not disclose, either directly or indirectly, the information to any other person or organisation other than members of their organisation who have been approved by the ABS to have individual access to the information
  • not attempt to match, with or without using identifiers, the data with any other list of persons or organisations
  • in relation to data made available via the RADL or the ABS Data Laboratory (ABSDL), access the data only in a manner specifically authorised in writing by the ABS; and
  • not attempt to access the data after the term of their authorisation is rescinded by the organisation which provided it, or after they cease to be a member of that organisation.

Use of the data for statistical purposes means use of the content of the CURF to produce information of a statistical nature, i.e. the arrangement and classification of numerical facts or data, including statistical analyses or statistical aggregates. Examples of statistical purposes are:

  • manipulation of the data to produce means, correlations or other descriptive or summary measures;
  • estimation of population characteristics;
  • use of data as input to mathematical models or for other types of analysis (e.g. factor analysis); and
  • providing graphical or pictorial representations of the characteristics of the population or subsets of the population.

All CURF users are required to read and abide by the Responsible Access to ABS Confidentialised Unit Record Files (CURFs) Training Manual available on the ABS Website (see Services We Provide/CURF Microdata/Accessing CURF Microdata).

Use of the data for unauthorised purposes may render the purchaser liable to severe penalties. Advice on the propriety of any particular intended use of the data is available from Micro Access Strategies Section via microdata.access@abs.gov.au

Conditions of sale

All ABS products and services are provided subject to the ABS conditions of sale. Any queries relating to these Conditions of Sale should be referred to intermediary.management@abs.gov.au

While the utmost care is taken in handling each CURF on CD-ROM, deterioration may occur between the time of copying and receipt of the file. Accordingly, if the CD-ROM is unreadable on receipt and this is reported to the ABS within 30 days of receipt, it will be replaced free of charge. 

Access method

Due to the level of detail provided, the 2006 TUS expanded CURF is only available via the ABS Remote Access Data Laboratory (RADL). The basic CURF is available on both CD-ROM and RADL. 

Price

The current recommended retail price of the TUS 2006 CURF is $1,320 (including GST) per CURF access type ($1,320 to access the basic CURF via CD-ROM and/or RADL, or $1,320 to access the expanded CURF via RADL). A bundled price of $1,980 (including GST) is available where clients request access to both the basic (whether on CD-ROM or RADL) and expanded CURFs in one single application. 

Accessing the CURF

All clients wishing to access the 2006 TUS basic and/or expanded CURF should refer to the ABS Website (see Services We Provide, CURF Microdata) and read the Responsible Access to ABS Confidentialised Unit Record Files (CURFs) Training Manual, and other relevant information, before downloading the appropriate Application and Undertaking forms and applying for access. 

Australian Universities:

University clients should refer to the ABS Website (see Services We Provide, Services to Universities). The 2006 TUS basic and expanded CURFs can be accessed by universities participating in the ABS/Universities Australia CURF Agreement for research and teaching purposes. 

Other clients:

Other prospective clients should contact the Microdata Access Strategies Section of the ABS at microdata.access@abs.gov.au or on (02) 6252 7714. 

Further information

For further information about accessing the CURF, clients should contact the Microdata Access Strategies Section of the ABS at microdata.access@abs.gov.au or on (02) 6252 7714. The CURF is not available on CD-ROM to overseas customers.

Data downloads

Data files

Survey material

To view the Time Use Survey 2006 Questionnaire (sample only) click here

To view the Time Use Survey 2006 Prompt Cards (sample only) click here

To view the Time Use Survey 2006 Diary (sample only) click here

Previous releases

 TableBuilder data seriesMicrodataDownloadDataLab
Time Use, 1997 Basic microdataDetailed microdata
Time Use, 1992 Basic microdata 

Appendix 1 - Activity classification

Show all

Appendix 2 - Data items

Show all

Appendix 3 - Survey instruments

Show all

Glossary

Show all

Abbreviations

Show all

Previous catalogue number

This release previously used catalogue number 4150.0.

Back to top of the page