A dissemination strategy designed to most effectively deliver the collection results to users will maximise the usefulness of survey or administrative data. This chapter examines key dissemination issues to consider including timeframe, methods, timeliness, access, confidentiality, data and metadata. This chapter also briefly discusses other dissemination issues such as revisions, handling requests for additional information and preventing misleading interpretations by users.
7.1 DISSEMINATION PLAN
A dissemination plan should outline how the collection results will be released. While developing the dissemination plan factors such as legislative obligations, needs of users, and confidentiality of data providers should be considered. The expected release date of the publication or data release should be made available to potential users well in advance. The dissemination plan should also address access issues such as what, when, how and who can access publications or releases through web, hardcopy or both.
As part of the dissemination strategy, you may need to develop a pricing policy taking into consideration the following issues:
· Public good requirements or community service obligations; and
· Whether your organisation charges for such publications and the basis for charging.
Even if agencies do not intend to disseminate their data or results (e.g. administrative data) it is a good idea to inform potential users if their data is accessible or available to others for sharing.
The Australian Bureau of Statistics now makes all its publications freely available on its website www.abs.gov.au
7.2 TIME TABLE FOR RELEASE OF STATISTICS
To maximise the relevance and usefulness, results and outputs should be released as soon as practicable after the collection. Issues to consider while developing the release time table include the time required to:
· check, validate and quality assure data and results;
· gain appropriate approval for the release of results;
· document systems and processes used in the collection, processing and analysis; and
· notify users in advance of the publication’s release date
7.3 FORMATS FOR DISSEMINATION
Data may be released in a number of formats. These comprise:
· Printed hardcopy publication;
· Website based publication;
· Other electronic publications (CD, electronic data files);
· Online access to registered, subscribed or general users;
· Data cubes;
· Briefing, seminars, conferences, media release and presentations;
· Telephone, FAX or email responses to special data requests;
· Customised products and services prepared on-demand for specific customers;
· Unit record files made available to authorised users through data centres and data laboratories
The information may be released using a combination of different formats and media depending on factors such as user needs, complexity and cost.
The timeliness of the publication refers to the time lag between the end of the reference period for the collection and the publication date. If a range of statistical outputs is to be released, broad or preliminary statistics can be released first, followed by more detailed releases based on more complete and detailed analyses. If there is a requirement for release before all planned processing or analysis is undertaken, then preliminary publications can be released with qualifying statements regarding the quality of the data.
You should formulate guidelines governing access of your publications arising from statistical collections based on your agency’s policies on dissemination. The ABS allows free and open access to its publications through its website. However, other agencies may their data and publications at a cost.
As discussed before, many administrative collections are undertaken by agencies for their own specific purpose and consequently these agencies may have strict policies and regulations on the access of such data. Some administrative collections are authorised by legislation and specifically preclude the release of certain information. Adherence to such legislative or government directives must always be followed, but this should not discourage the dissemination of statistical outputs from these collections to other users where possible. For example aggregated data tables may be produced and released for such data even though it would not be possible to release the unit record data.
Data confidentiality is the assurance that information about individual respondents is not released outside the organisation and cannot be derived from published results. This should be supported by the dissemination process. See Chapter 11- Confidentiality and Privacy for further information on confidentiality and privacy issues you need to be aware of while handling statistical collections.
7.7 QUALITY DECLARATION
All statistical publications should be accompanied by a statement on the quality of data to help in the appropriate use of the statistics and to help users make informed decisions. The Data Quality Declaration should include an assessment of the data against each of the seven data quality framework criteria outlined in section 2.5 in this Handbook.
7.8 METADATA AND EXPLANATORY NOTES
Metadata is information about the actual data. Metadata includes information such as concepts used in the collection, classifications, scope and coverage, sampling methodology, units and measurements, derivations and transformations, estimations and adjustments and processing techniques (e.g. editing).
An explanatory note containing key aspects of the metadata will help the users understand the data as well as concepts and processes used in the collection cycle. The metadata and explanatory notes also serve as an important document for use in future similar statistical collections.
See Chapter 9 – Data Management for further information on metadata..
Some published data may need to be revised and republished especially when new information becomes available which can significantly alter or add value to the original information. Revisions are also necessary when there is a need to release preliminary estimates before the data is fully analysed. A revision policy should balance the need for the best estimates against the uncertainty created by frequent changes and the cost of revisions and their publication. The explanatory notes accompanying the main results should include a section on any planned revisions.
7.10 USER NEEDS
Agencies undertaking data collections should have adequate systems and procedures to meet the various information requests from different users. Users should be provided with contact information for following-up such as the person, work area, telephone number and email address. These should be available for queries or feedback or requirements for further information regarding the publication. Specialist users may also want information such as the availability of more detailed datasets, and tables, data cubes and unit records.
When dealing with these requests agencies should consider:
· the confidentiality of the data requested;
· the agency’s pricing policy for such requests and the price to charge;
· the time and resources required to fulfil such requests; and
· any conditions and restrictions to be followed by clients for use of the data.
7.11 MISUSE OF STATISTICS
Statistics are often misunderstood and are prone to misuse and misinterpretations. While it is not possible to completely prevent misuse, it is advisable for to keep a vigil on the proper use of the statistics produced. Widespread misuse or misinterpretation may damage the reputation and credibility of the agency which collected the statistics. It is advisable for agencies to have procedures to handle misuse of their statistics by the media, businesses or individuals. Agencies can also educate users on the proper interpretation of statistics as part of the public relationship management strategy.