Australian Bureau of Statistics
4363.0.55.001 - Australian Health Survey: Users' Guide, 2011-13
Latest ISSUE Released at 11:30 AM (CANBERRA TIME) 19/07/2013
|Page tools: Print Page Print All RSS Search this Product|
At least nine days after the initial face to face interview respondents were asked to participate in a Computer Assisted Telephone Interview (CATI). This interview collected the data recorded in the pedometer diary, along with the 2nd day 24 hour dietary recall.
Determining valid days reported
To be retained on the file and considered a valid day, the reported step count was required to be six or more steps. This is due to the model of pedometer used not recording steps until there are at least six continuous incidental steps registered. Aside from this, there were no maximum or minimum cut-offs applied to the number of step counts in the detailed data or calculations because the nature of the self-reported information around why the pedometer was removed was insufficient to determine whether or not the step count was representative of a person’s level of physical activity.
Respondents were requested to wear the pedometer for eight days. 18% of respondents who participated in the pedometer component did not record a valid step count for the full 8 days (a total of 6061 respondents participated in the pedometer component).
Data was collected for any number of days up to 8 days and was not always 8 consecutive days. As a result, respondents may have duplicate days of the week reported, particularly if they completed 8 days (where at least one day had to be the same day of the week). The following table demonstrates the number of pedometer records per day and the number of respondents reporting per day.
Minimum Day threshold
While all days with reported steps have been retained on the Pedometer level data file, to be included in certain summary data items (i.e. averages as outlined in the Data item list) on the person level a minimum days threshold was applied. This required a respondent to report at least four days of pedometer data, including at minimum one week and one weekend day.
This resulted in 7% of respondents who participated in the pedometer component being excluded from the majority of person level derivations with the application of this threshold and, where applicable, they have been placed into a category called ‘Did not meet day threshold requirement’. A data item ‘Pedometer day threshold requirement’ has also been produced for use with these items if required.
Time worn data
Total minutes wore the pedometer that day was calculated by subtracting the time the pedometer was first put on from the last time the pedometer was removed that day, minus any time the respondent removed the pedometer during the day. If any of these components were unknown, then this field was set to ‘Not known’ (approximately 0.2% of days recorded).
Total minutes took the pedometer off that day was calculated by aggregating all the minutes from each time the pedometer was taken off. If any time was not known, then the total time was set to ‘Not known’.
There were no minimum threshold for the number of minutes a respondent wore the pedometer during a given day, however, there was a maximum time of twenty four hours.
The data was examined by looking at outliers for step counts, incremental step counts (pedometer not reset), non-consecutive and erroneous date values and incorrect recording of 12 hour time instead of 24 hour time. Where it was determined that these items were incorrectly recorded, the data was amended to reflect either the correct values or set to a ’not known’ code. Days with less than six or unknown step counts were removed from the data file.
A number of respondents recorded the time the pedometer was put on as 00:00 and the time the pedometer was last taken off as 00:00. In these cases, the record for the last time taken off was amended to 23.59 in order to meet time derivation requirements.
This time data is available for use, should users wish to implement thresholds in their calculations. Time data has not been used in the derived data items produced for this survey.
Removal of Pedometer
There are certain physical activities where a pedometer may not be worn (such as swimming, cycling or contact sports). Removal of the pedometer for these activities could therefore result in an underestimate of the steps taken. Respondents were asked to record the reason and duration of pedometer removal during the day. Of the child pedometer days recorded, 8% had noted the removal for sport or swimming, while for adults, just 3% of pedometer records specified removal for exercise or swimming. There was no imputation of steps for any time the pedometer was removed regardless of the reason for removal. As per determining a valid day, there was insufficient detail collected in the daily activity sheet to determine a level or intensity of any physical activity that was reported (e.g. swimming) as the reason for removal. ‘Reason for removal’ has therefore not been made available as a data item on the file.
Pedometer step thresholds
There are no recognised national guidelines which indicate a minimum number of steps per day required for adults or children to maintain a healthy lifestyle. Several thresholds have been used in a number of publications to show the proportion of the population who reached a desired level of physical activity. The following table indicates the thresholds for the number of step counts that have influenced the data items produced or data reported on for this survey.
In addition, data items based on average pedometer steps have been produced that utilise pedometer-determined physical activity measures proposed for Adults and Children (6-12 years). In order to provide a measure for the full pedometer participant population, 5 year olds have been included with the 6-12 year olds and children 13-17 years have had cut-offs applied between the adults and the younger children cut-offs.
Step count data has been assigned the NNPAS person level weights. As such, population estimates for participating respondents published from this survey will not match the total population. Analysis was undertaken on the responding population and they are considered representative of the population generally. Therefore, if required, population estimates may be produced using the proportion results converted to an estimate using the total for the desired population. For more details regarding weighting and analysis undertaken, see Weightings, Benchmarks and Estimation procedures in this Users’ Guide.
The data items and related output categories for this topic are available in Excel spreadsheet format from the Downloads page of this product.
Points to be considered when interpreting data for this topic include the following:
Comparability with other surveys
The NNPAS is the first ABS survey to collect step counts using a pedometer.
Pedometer data has been collected in other non-ABS surveys. However, caution must be taken when interpreting results due to differences in scope, collection methodology, the device used to measure the number of steps, and any thresholds applied in the final analysis.
This page last updated 10 December 2013
Unless otherwise noted, content on this website is licensed under a Creative Commons Attribution 2.5 Australia Licence together with any terms, conditions and exclusions as set out in the website Copyright notice. For permission to do anything beyond the scope of this licence and copyright terms contact us.