After American Time Use Survey (ATUS) raw data are collected and coded, they go through an editing and imputation process that allows them to be used to produce estimates of time spent in daily activities. Many of the edits performed on ATUS data deal with item nonresponse. Imputation is a common way of treating item nonresponse, and many ATUS variables are imputed if missing. Chapter 6 of the User’s Guide specifies the different imputation methods used for different variables, and the ATUS data dictionaries have further information on how to identify edited variables and their allocated values. In addition to the various imputation procedures, data edits are performed for confidentiality (any data elements identifying the respondent are removed from the microdata files) and consistency, some variables are recoded, and the data are weighted. ATUS weights are designed to reduce any bias in the estimates that is due to differences in sampling and response rates across subpopulations and days of the week. For more information about the weights, see chapter 7 of the User’s Guide. Once the data have been through processing, editing, and imputation, the edited data sets are ready for analysis and eventual publication. Annual and multiyear ATUS microdata files are available for free from the ATUS website at www.bls.gov/tus/data.htm.
Chapter 7 of the ATUS User’s Guide contains information about the types of estimates that can be generated with the ATUS data and the formulas needed to produce those estimates. Numerous types of estimates and analyses can be produced with the ATUS data; however, there are three main types of ATUS estimates: average hours, participation rates, and number of people. ATUS estimates are representative of the civilian noninstitutional population ages 15 and older and of various subpopulations. Official ATUS estimates are published annually.
To generate estimates, researchers and others may use ATUS microdata files together with a statistical software package, such as SAS, Stata, or SPSS. Researchers use the data in a variety of ways. For example, one can examine the time use of particular subpopulations, look at time use patterns throughout the course of the day, investigate trends in time use over the course of several years, or explore particular activities or combinations of activities. ATUS microdata files also can be linked with CPS data files for further research.
ATUS estimates are available as time series from the ATUS data base at https://www.bls.gov/tus/labstattips.htm.
Statistics based on the ATUS are subject to both sampling error and nonsampling error. When a sample, rather than the entire population, is surveyed, estimates differ from the true population values they represent. The component of this difference that occurs because samples differ by chance is known as sampling error, and its variability is measured by the standard error of the estimate.
Sample estimates from a given survey design are unbiased when an average of the estimates from all possible samples would yield, hypothetically, the true population value. In this case, the sample estimate and its standard error can be used to construct approximate confidence intervals, or ranges of values that include the true population value accompanied by known probabilities. If the process of selecting a sample from the population were repeated many times, an estimate made from each sample, and a suitable estimate of its standard error calculated for each sample, then approximately 90 percent of the intervals from 1.645 standard errors below the estimate to 1.645 standard errors above the estimate would include the true population value.
The ATUS data also are affected by nonsampling error, which is the average difference between population and sample values for samples generated by a given process. Nonsampling error can occur for many reasons, including the failure to sample a segment of the population, inability to obtain information from all respondents in the sample, inability or unwillingness of respondents to provide correct information, and errors made in the collection or processing of the data. Errors can also occur if nonresponse is correlated with time use. The average annual response rates for the ATUS can be found in chapter 3.6 of the ATUS User’s Guide. The ATUS statistical weights in part adjust for nonsampling error, and quality assurance procedures are used to minimize nonsampling, data entry, and coding errors in the survey estimates.
Although attempts have been made to collect the most accurate data possible, the ATUS data do have limitations. For example, with the exception of childcare, information on secondary activities (activities that are done at the same time as the primary activity) is not collected. This decision not to collect such information could lead to underestimates of the amount of time people spend doing activities that are frequently done in combination with other activities. For instance, ATUS estimates likely underestimate the amount of time people spend listening to music, because so many people listen to music while doing other things.
Estimates appearing in ATUS tables must meet reliability standards before being presented to the public. In 2010, a new standard was developed that takes into account the coefficient of variation, standard error, and number of observations available before reporting an estimate. Prior to 2010, a standard was in place that included only the sample size or population base for the estimate.
Beginning with 2010 data, ATUS estimates of average hours per day and participation rates are not published unless there are a minimum number of respondents representing the given population. Additional publication criteria are applied that include the number of respondents who reported doing a specified activity and the standard error or coefficient of variation for the estimate. Estimates that are considered “close to zero” or that round to zero (e.g., 0.00 for estimates of hours) are published as approximately zero, or “~0.”
Chapter 10 of the User’s Guide has additional information on ATUS data quality. For a detailed description of the statistical reliability criteria necessary for publication, please email the ATUS staff at ATUSinfo@bls.gov.
Last Modified Date: March 20, 2017