8155.0 - Australian Industry, 2000-01  
ARCHIVED ISSUE Released at 11:30 AM (CANBERRA TIME) 27/08/2003   
   Page tools: Print Print Page Print all pages in this productPrint All

LIMITATIONS OF FINANCIAL DATA ANALYSIS

RELATIVE STANDARD ERROR

1 Since the estimates in this publication are based on information obtained from a sample drawn from units in the surveyed population, the estimates are subject to sampling variability. That is, they may differ from the figures that would have been produced if all units had been included in the survey.

2 One measure of the likely difference is given by the standard error (SE), which indicates the extent to which an estimate might have varied by chance because only a sample of units was included. The relative standard error (RSE) provides an immediate indication of the percentage errors likely to have occurred due to sampling, and thus avoids the need to refer to the size of the estimate. Tables 3.1 and tables 3.6, 3.7 and 3.8, provide RSEs for a selection of estimates presented in this publication.

3 The relatively small sample size of the EAS (directly collected) collection does not allow for the compilation of reliable estimates generally below the ANZSIC subdivision level. However, by utilising the very large tax based sample it is possible to produce ANZSIC class level estimates. One of the measures of quality of the class level estimates are the relative standard errors (RSEs) contained in table 3.1. Approximately 88% of the ANZSIC class level estimates have RSEs of 25% or less. Some of the RSEs, are relatively large and therefore the estimates to which they relate should be used with extreme caution.

4 There are about two chances in three that the difference between the estimate shown and the true value will be within one SE, and about 19 chances in 20 that the difference will be within two SEs. Thus, for example, if the estimated value of a variable is $12,000m and its RSE is 5%, its quality in terms of sampling error can be interpreted as follows. There are about two chances in three that the true value of the variable lies within the range $11,400m to $12,600m, and 19 chances in 20 that it lies within the range $10,800m and $13,200m.

5 The size of the RSE may be a misleading indicator of the quality of some of the estimates for Operating profit before tax (OPBT). This situation may occur where an estimate may legitimately include positive and negative values reflecting the financial positions of different business entities. In these cases the aggregate estimate can be small relative to the contribution of individual business entities, resulting in a SE which is large relative to the estimate.

6 The EASTAX sample is not selected on the basis of state and this could have an impact on the size of the sampling error at the state level. To some extent this is offset by the use of business income tax data which increases the sample size, resulting in a broader coverage of units for each state.

NON-SAMPLING ERROR

7 The imprecision due to sampling variability, which is measured by the SE, is not to be confused with inaccuracies that may occur because of 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. Inaccuracies of this kind are collectively referred to as non-sampling error and they may occur in any enumeration, whether it be a full count or a sample.

8 While it is not possible to quantify non-sampling error, every effort is made to reduce it to a minimum. Collection forms are designed to be easy to complete and assist businesses to report accurately. Efficient and effective operating procedures and systems are used to compile the statistics.

9 There are also non-sampling errors associated with the ATO business income tax file. For example, the ATO accounts for non-response in the business income tax file by either bringing forward the previous year's data for a non-responding business, or leaving the data as zero if the business does not have an ATO response history.

10 Differences in accounting policy and practices across businesses and industries can also lead to some inconsistencies in the data used to compile the estimates. While much of the accounting process is subject to standards, there remains a great deal of flexibility available to businesses in the accounting policies and practices they adopt.

11 Chapter 3: Experimental estimates, contains experimental estimates at the ANZSIC class level. This is the finest level of classification in the ANZSIC. It is only the incorporation of ATO business income tax data that has made it feasible to produce estimates at this level of detail.

12 The class level estimates in this publication can sometimes be different to the class level estimates produced by the Service Industries program of surveys which focus on fine level detail for a given ANZSIC class. In most cases the differences are due to differences in scope and coverage. However, differences can occur due to inconsistencies in ANZSIC coding on the ABS Register of Businesses, ANZSIC coding on the ATO Business Income Tax file and ANZSIC coding undertaken as part of the Service Industries program.

13 Chapter 3 also contains state experimental estimates. Users should be aware that because direct collection has not been used to apportion EASTAX estimates to states, some non-sampling error will result from the techniques used. For full details of the methodology used to allocate estimates to states please refer to the Technical Note 1: Methodology.

14 The above limitations are not meant to imply that analysis based on this data should be avoided, only that they should be borne in mind when interpreting the data presented in this publication.

15 This publication presents a wide range of data that can be used to analyse business and industry performance.

16 It is important that any analysis be based upon the range of data presented rather than focusing on one variable.