GlossaryStatistics and interpretation
Data suppression
Data suppression withholds selected values or records under a stated rule. Learn why suppressed cells differ from zero, unknown, and missing values.
Data suppression is the withholding of selected values or records from a release under a stated rule. It can protect confidentiality, limit misleading interpretation of unreliable statistics, or meet another publication requirement. A suppressed value is unavailable to the reader; it should not be treated as zero or assumed to have a particular size.
What the symbol means
A table may replace a value with a letter, dash, or other symbol. The legend and methods explain whether it means suppression, missing information, or something else. The same symbol can have different meanings in different products.
The Census Bureau describes primary suppression of sensitive cells and complementary suppression of additional cells that could otherwise reveal the protected value. CDC's U.S. Cancer Statistics also explains suppression for confidentiality and reliability. Its threshold belongs to that specific product and is not a universal public-safety rule. Census disclosure guidance; CDC suppression notes.
Example
Hypothetical example: An agency publishes a table of sensitive service categories. One category is marked “suppressed,” and the note explains that selected values are withheld for privacy. The remaining visible categories sum to 90.
The analyst can report the visible subtotal as 90, but cannot label it the complete total unless the publication establishes that it is complete. Replacing the suppressed entry with zero would falsely assert that no qualifying records existed. The agency may also withhold another entry or total to prevent deduction of the protected value.
Common confusion
Suppression is different from a genuinely observed zero. It is also different from an unknown value, a field that does not apply, or a category the source never collected. These can all produce blank-looking cells while requiring different interpretations.
A suppressed cell does not necessarily contain a small count. Complementary suppression can hide another value to protect the initially sensitive cell. A reader should not infer an exact range unless the publisher expressly defines one.
Preserve the publication limit
Carry suppression labels into summaries and downloads rather than allowing software to turn them into zeros. Explain when totals include withheld components that are absent from the visible detail.
Combining periods or categories may permit a more useful public summary, but any release decision needs the source's privacy and reliability rules. Do not reconstruct intentionally protected values by subtracting visible entries from totals. A clear report states what remains visible and what conclusions the withheld information prevents.