EXPLANATORY NOTES

INTRODUCTION

1 This release presents results from the Internet Activity Survey (IAS) conducted in respect of the three months ended 31 December 2016.


SUBSCRIBERS

2 The ABS counts the number of subscribers that have an internet connection with an Internet Service Provider (ISP) on the last day of the reference period. ABS subscriber statistics measure the number of 'subscriber lines' rather than the number of 'users' and therefore, counts of subscribers are not the same as counts of people/organisations with internet access. This is because some subscribers may have accounts with more than one ISP or multiple accounts with a single ISP. Conversely, there are single ISP subscriber accounts that provide internet access for multiple people/organisations (e.g. universities).


STATISTICAL UNIT

3 The unit for which statistics are reported in the IAS is the legal entity providing internet access.

4 ISPs in the IAS have been classified by size according to the number of subscribers that ISPs reported at the end of the reference period. The size categories are defined as follows:

Size of ISP	Number of subscribers
Very small ISP	1 - 100
Small ISP	101 - 1,000
Medium ISP	1,001 - 10,000
Large ISP	10,001 - 100,000
Very large ISP	100,001 or more

SCOPE AND COVERAGE

5 The December 2016 IAS covers ISPs who operated in Australia as at 31 December 2016 that had more than 1,000 subscribers. ISPs are defined as businesses that supply internet connectivity and access services to individuals, households, businesses, government and other organisations. Hotels, libraries, internet kiosks, internet cafes and hotspots that provide internet access on a casual basis are excluded from the survey population.

6 The source of the IAS population frame is a list of ISPs from the Telecommunications Industry Ombudsman (TIO), with whom ISPs are required to register. Using this list as a basis, the ABS maintains its own list of ISPs. The TIO register may contain the following organisations that are out of scope or counted differently on the IAS: those that are no longer operating as ISPs, including wholesalers and web hosting ISPs only; those where there is common ownership and, therefore reporting is consolidated for these ISPs on the IAS; and other types of organisations that are required to register with the TIO (e.g. telecommunications companies that do not provide ISP services).

7 The reference periods for IAS are June and December to align with international reporting of broadband statistics by the Organisation for Economic Co-operation and Development (OECD). The IAS includes information reported by all ISPs operating in Australia, with more than 1,000 subscribers, as at 30 June and 31 December.

8 Internet connections via a mobile handset were first published in June 2010 as an appendix and are now part of the main publication.


REFERENCE PERIOD

9 This release relates to information as at 31 December 2016, except for volume of data downloaded, which is for the three months ended 31 December 2016.


IMPUTATION

10 Where data were missing or required clarification, respondents were contacted in the first instance. When necessary, missing data were imputed based on historical data.


ROUNDING

11 Where figures have been rounded, discrepancies may occur between the sum of components and the total.


RELIABILITY OF DATA

12 As the IAS does not have a sample component, the data are not subject to sampling variability. However, other inaccuracies, collectively referred to as non-sampling errors, may affect the data. These non-sampling errors may arise from a number of sources, including:

• errors in the reporting of data by respondents;
• errors in capturing or processing of data;
• estimation for missing or misreported data; and
• definition and classification errors.

13 Every effort has been made to reduce non-sampling error to a minimum by careful design and testing of questionnaires, efficient operating procedures and systems, appropriate methodology, and contact with providers to resolve anomalies. Errors in previous cycles' data are only revised (if possible) where they have a significant impact on the data released.

14 The statistics on volume of data downloaded should only be considered as an indicative measure of internet activity during the reference period and therefore should be used with caution. The ability of ISPs to report volumes of data downloaded is variable. For example, there may be differences in reporting of billable vs. free content, downloads and uploads may not always be separable and ISPs may use recording systems that are not extracting data download information on an equivalent basis.


ACKNOWLEDGEMENT

15 The ABS releases draw extensively on information provided freely by individuals, businesses, governments and other organisations. Their continued cooperation is very much appreciated for without it, the wide range of statistics published by the ABS would not be available. Information received by the ABS is treated in strict confidence as required by the Census and Statistics Act 1905.


RELATED RELEASES

16 Other ABS releases on information technology and telecommunications in Australia are:

Business Use of Information Technology, cat. no. 8129.0
Household Use of Information Technology, Australia, cat. no. 8146.0
Summary of IT Use and Innovation in Australian Business, cat. no. 8166.0
Selected Characteristics of Australian Business, cat. no. 8167.0