"Dissemination means agency initiated or sponsored distribution of information to the public. Dissemination does not include distribution limited to government employees or agency contractors or grantees; intra- or inter-agency use or sharing of government information; and responses to requests for agency records under the Freedom of Information Act, the Privacy Act, the Federal Advisory Committee Act, or other similar law. This definition also does not include distribution limited to correspondence with individuals or persons, press releases, archival records, public filings, subpoenas or adjudicative processes." - OMB Guidelines.
The first key point in disseminating statistical information is the principle of openness relative to all aspects of quality. Pursuant to that principle, along with the statistical information being disseminated, the related supporting documentation must also be available. That documentation can be separately posted documents referenced by the disseminated information or it can be part of the disseminated entity. The second key point in the dissemination is the final reviews before dissemination. These quality reviews are a final assurance that all quality control steps have been taken and that the dissemination package is complete.
5.1 Publications and Disseminated Summaries of Data
- In publications or summaries, information should be clearly presented to users, and users should be informed about the source(s) of the information presented.
- As far as possible, tables, graphs, and figures should be interpretable as standalone products in case they become separated from their original context.
- Methods used to produce data displayed in tables, graphs, and summary data should be available to the reader.
- Statistical interpretations should indicate the amount of uncertainty.
- Documents should be well organized with language that clearly conveys the message intended. Tables, graphs, and figures should be consistent with each other and the text discussing them.
- All tables, graphs, figures that illustrate data, and text that provides data not in accompanying illustrations should include a title. Titles for tables and graphs should be clearly worded and answer three questions: what (data presented), where (geographic area represented), and when (date covered by data).
- All tables, graphs, figures that illustrate data, and text that provides data not in accompanying illustrations should include a source reference. The source reference should contain one or more entries with references to the sources for the information presented. The reference should be sufficiently detailed for a reader to locate the data used. Since databases and documents may be updated, the "as of" date for the source should also be noted.
- Footnotes should be used, if necessary, to clarify data illustrations, tables, graphs, and figures to clarify particular points, abbreviation symbols, and general notes.
- The style of a publication should conform to specific agency style guidelines to ensure consistency and clarity throughout the document.
- Documents disseminated on the Internet must be accessible as required by section 508 of the Rehabilitation Act (29 USC 794d).
- A contact point should be provided in the publication or with the summaries to facilitate user comments and suggestions.
- Documents containing estimates, projections, and analyses should contain or reference the methodology supporting documentation required in sections 4.3 and 4.4.
- U.S. Government Printing Office Style Manual
5.2 Micro Data Releases
- The term "micro data" refers to data files with various information at the "unit" level. The unit is dependent upon what data are being collected and from what sources.
Examples: micro data may be a collection of individual responses from each person or each household to a survey, reports of information from each company, or reports of individual incidents.
- Making micro data available can enhance the usefulness of the information, and can assist the public in determining whether results are reproducible. However, micro data should not be released in violation of existing protections of privacy, proprietary information, or confidentiality.
- Micro data should be provided in a manner that facilitates its usefulness to users.
- Quality information as recommended herein, file layouts, and information describing the data (i.e., metadata) enhance the usefulness of the micro data.
- Micro data released to the public should accessible by users with generally available software. It should not be restricted to a single application format.
- Micro data should be accompanied by (or have a reference to) the quality-related documentation discussed herein: planning documentation and collection, processing, and analysis methodology.
- Micro data releases should be accompanied by file layouts and information describing the data.
- Micro data should be accompanied by a clear description of revision information related to the file.
- A contact point should be provided with the data to facilitate user comments and suggestions.
- International Standardization Organization standard 11179, Specification and Standardization of Data Elements.
5.3 Source and Accuracy Statements
- Source and Accuracy Statements (S&As) are compilations of data quality information discussed herein. They provide information on where the data came from, how it was collected, and how it was processed. They include information on known strengths and weaknesses of the data.
- S&As should be regularly updated to include changes in methodology and results of any quality assessment studies.
- The S&A for a data source should contain or refer to the current data system objectives and data requirements as discussed in sections 2.1 and 2.2 of these guidelines.
- The S&A for a data source should contain the data source and data collection design as discussed in sections 2.3 and 2.4 of these guidelines.
- The S&A for a data source should contain or refer to the collection operations methodology documentation discussed in sections 3.2 and 3.3.
- The S&A for a data source should contain or refer to the processing documentation discussed in sections 4.1 - 4.4.
- The S&A for a data source should describe major sources of error including coverage of the target population, missing data information, measurement error, and error measures from processing quality assurance.
- The S&A for a data source should contain or reference the revision process for the system and should indicate the source for revision information for the data source.
- General Accounting Office, Performance Plans: Selected Approaches for Verification and Validation of Agency Performance Information, GAO/GGD-99-139 (July 1999).
- Office of Management and Budget, Statistical Policy Working Paper 31: Measuring and Reporting Sources of Error in Surveys (July 2001).
5.4 Pre-Dissemination Reviews
- Informal and formal reviews of publications, summaries, or micro data will help ensure that a data product meets a minimal level of quality.
- Due to the diverse aspects of quality in a final product, reviews need to be conducted by several people with different backgrounds.
- Reviews of documentation produced through the various stages of data development will enhance the review process.
- A subject matter specialist other than those directly involved in the data collection and analysis should review the plans, methodology documents, and reports prior to dissemination. They should also review publications and summaries resulting from the data for content and consistency.
- Publications should be reviewed by a style and visual information specialist for compliance with style standards.
- A statistician or other data analysis specialist other than those directly involved in the data collection and analysis should review the plans, methodology documents, and reports prior to dissemination for compliance with these guidelines. They should also review publications and summaries resulting from the data for the wording of statistical interpretation.
- Any items to be disseminated via the Internet must be reviewed by a Section 508 compliance specialist for accessibility.
- Any data products that will be disseminated via special software onto the Internet should be tested for accessibility and interpretability prior to dissemination.
- For micro data releases, the release files and the metadata should be reviewed by an information technology specialist for clarity and completeness.
- If an external peer review process is used: (1) peer reviewers should be selected primarily on the basis of necessary technical expertise; (2) peer reviewers should be expected to disclose to DOT prior technical/policy positions they may have taken on the issues at hand and their sources of personal and institutional funding (private or public); and (3) peer reviews be conducted in an open and rigorous manner.
- Ott, E., E. Shilling, and D. Neubauer. 2000. Process Quality Control: Troubleshooting and Interpretation of Data. New York, NY: McGraw-Hill.