Disclosure Review Policies and Procedures
Our disclosure review policies and procedures exist to protect the confidentiality of NCHS study participants. Please read the Disclosure Manual: Rules for Researchers.
General Output Policies
- Data sets will not be released.
- Output in the form of a data set will not be released.
- Absolutely NO output will leave the RDC facilities without first being reviewed by an RDC Analyst or the remote access system.
- Output MUST match the research questions/output suggested in the proposal.
- Cell Suppression Criteria Employed by the RDC: Guidelines may differ by data system and possibly by survey year because of sample size, sample design, and content. Sometimes specific projects pose additional concerns and will experience additional or more stringent suppression.
- Approved output is returned via email (within hours for ANDRE users, within 5 days for onsite users, and within 3 weeks for Census users).
- Although the output files are reviewed by an RDC staff member or the remote access system for disclosure concerns, it is your responsibility to use these statistics in a way that doesn't pose additional risk to the respondents.
On-Site Users: Preparing Your Output for Review
- Make sure output submitted for review includes:
- What is this output? (This can be a title: a regression of…)
- Who is the (sub)sample in this analysis? (black males age 20-29)
- How will output be used? (publication, presentation, etc.)
- Before submitting output for review by your RDC Analyst, you must review your output for the following:
- Make sure it is in a form that can be released by the RDC. Output will be released in the default format produced by the statistical package (i.e. .lst files for SAS) or in a human-readable plain text file (i.e., files that can be opened and that are readable in notepad such as tab delimited text .txt files or comma-separated values .csv files). We encourage you to populate the actual tables that will appear in your publication and submit those for review.
- If there is any output that you feel could lead to the identification of an individual or institution, remove it. If you have questions, discuss this with your RDC Analyst.
- Any procedure that produces output on an individual or institution must be removed.
- Printing individual cases is not permitted. Remove any individual level data from your output.
- Extreme values or values representing an individual must be removed. Examples include minima, maxima, medians, and modes. If a procedure, such as Proc Univariate creates extreme observations, 0, 1, 99, and 100 percentiles, those extreme values must also be removed.
- If you do not review your output before requesting that your RDC Analyst review it, you may be asked to return to the RDC to do so.
- All cells with a frequency less than 5 should be asterisked before they are submitted to your RDC Analyst. If at all possible, we encourage you to consider re-categorizing the variables instead of requesting suppression.
- You must make a special request (which will take time to review) if you wish to have intermediate output (output that will not appear in publication) reviewed. This output can certainly be created and used on site, but we discourage its release.