TECHNICAL NOTE DATA QUALITY


Reliability

Quality Indicators


RELIABILITY

1 The estimates in this release are based on information obtained from a sample survey conducted by the Australian Bureau of Statistics (ABS). The reliability of statistics derived from any collection may be affected by a range of factors that are independent of the methodology used. These factors result in non-sampling error. Sample surveys are also subject to inaccuracies that arise from the fact that a sample was selected rather than conducting a census. This type of error is called sampling error.


Sampling error

2 Sampling variability occurs when a sample, rather than the entire population, is surveyed. It reflects the difference between estimates based on a sample and those that would have been obtained had a census been conducted. One measure of the likely difference is given by the standard error, which indicates the extent to which an estimate might have varied by chance because the data were obtained from only a sample of units. There are about two chances in three that a sample estimate will differ by less than one standard error from the figure that would have been obtained if all businesses/organisations had been included in the survey, and about 19 chances in 20 that the difference will be less than two standard errors.

3 Another measure of sampling variability is the relative standard error (RSE), which is obtained by expressing the standard error as a percentage of the estimate to which it refers. The RSE is a useful measure in that it provides an immediate indication of the sampling error in percentage terms, and this avoids the need to refer also to the size of the estimate. RSEs for earnings for each industry are shown in the table below. Detailed RSEs can be made available on request.

RELATIVE STANDARD ERRORS
	Earnings
	%
Mining	5.3
Manufacturing	3.1
Electricity, gas, water and waste services	1.8
Construction	6.3
Wholesale trade	6.2
Retail trade	6.6
Accommodation and food services	14.8
Transport, postal and warehousing	6.0
Information media and telecommunications	7.2
Financial and insurance services	4.8
Rental, hiring and real estate services	9.8
Professional, scientific and technical services	6.3
Administrative and support services	10.7
Public administration and safety	0.7
Education and training	2.4
Health care and social assistance	2.6
Arts and recreation services	6.3
Other services	5.5
All Industries	1.4

4 To illustrate, the estimate of earnings during 2010-11, for the Mining industry, was $16,102.3 million. The RSE of this estimate is shown as 5.3%, giving a standard error of approximately $853.4 million. Therefore, there are two chances in three that, if all units had been included in the survey, an estimate in the range of $15,248.9 million to $16,955.7 million would have been obtained. Similarly, it implies that there are nineteen chances in twenty (i.e. a confidence interval of 95%) that the estimate would have been within the range of $14,395.5 million to $17,809.1 million.

5 Estimates for all Major Labour Costs Surveys were annotated with:

• a single asterisk (*) if their RSE lay in the range 25% to less than 50%
• a double asterisk (**) if their RSE was 50% or more.

6 For the 2010-11 Major Labour Costs Survey, estimates were annotated with a carat (^) if their RSE lay in the range 10% to less than 25%. Estimates for previous surveys did not distinguish RSEs in this range. Users are advised to bear these differences in mind when comparing estimates across reference periods.


Non-sampling error

7 Error other than that due to sampling may occur in any type of collection, whether a full census or a sample, and is referred to as non-sampling error. All data presented in this publication are subject to non-sampling error. It can arise from inadequacies in available sources from which the population frame was compiled, imperfections in reporting by providers, errors made in collection such as in recording and coding data and errors made in processing data. It also occurs when information cannot be obtained from all businesses/organisations selected. The imprecision due to non-sampling variability cannot be quantified and should not be confused with sampling variability, which is measured by the standard error.

8 Although it is not possible to quantify non-sampling error, every effort was made to reduce it to a minimum. Collection forms were designed to be easy to complete and assist businesses/organisations to report accurately. Efficient and effective operating procedures and systems were used to compile the statistics. The ABS compared data from different ABS (and non-ABS) sources to ensure consistency and coherence.

9 Differences in accounting policy and practices across businesses/organisations can lead to some inconsistencies in the data used to compile the estimates. Although much of the accounting process is subject to standards, there remains a great deal of flexibility available to individual businesses/organisations in the accounting policies and practices that they adopt.

10 The above limitations are not meant to imply that analysis based on these data should be avoided, only that the limitations should be borne in mind when interpreting the data presented in this publication. This publication presents a wide range of data that can be used to analyse business/organisation performance. It is important that any analysis be based upon the range of data presented rather than focusing on one variable.


QUALITY INDICATORS

11 In the 2010-11 Major Labour Costs Survey, there was an 89.5% response rate from all businesses/organisations that were surveyed and found to be operating during the reference period. Data were imputed for the remaining 10.5% of operating businesses/organisations. Imputed responses contributed 10.9% to the estimate of total labour costs for all selected industries.