SAHIE Methodology - Census Bureau

Source and Accuracy

The data provided are indirect estimates produced by statistical model-based methods using sample survey, decennial census, and administrative data sources. To maintain confidentiality, the Census Bureau uses procedures to assure that the estimates and related information that are released cannot be used to disclose individual data or violate other confidentiality restrictions applicable to the source data. The estimates contain error stemming from model error, sampling error, and nonsampling error. Confidence intervals are provided to indicate the quality of the estimates. Subject to the validity of the underlying model assumptions, these reflect uncertainty due to the effects of model error and sampling error, but do not account for the effects of nonsampling error. In addition, one should exercise caution when making comparisons between these model-based estimates and other Census Bureau estimates. See methodology for full technical documentation of each set of estimates produced. See information about data inputs for information on the data sources used in our models.

Error in model-based estimates

Error in model-based estimates arises from the effects of model error, sampling error, and nonsampling error. The relative contribution of these error components to the error in the model-based estimates depends on the model used and the properties of the data. Standard errors provided for the model-based estimates reflect, to the extent possible and subject to the model assumptions, the contributions of model error and sampling error, but do not reflect the contribution of nonsampling error.

Model error

Model error refers to error that would result in predictions from a statistical model even with no errors in the data (no sampling or nonsampling error). It can generally be broken down into three components: error that would occur in the predictions even if the true model were known; error resulting from estimation of model parameters; and error resulting from differences between the form of the assumed model and that of the true, unknown, model.

Sampling error

Sampling error is the difference between an estimate based on a sample and the corresponding value that would be obtained if the estimate were based on the entire population (as from a census). Note that sample-based estimates will vary depending on the particular sample selected from the population. Measures of the magnitude of sampling error in direct survey tabular estimates (variances, standard deviations, or coefficients of variation) reflect the variation in the estimates over all possible samples that could have been selected from the population using the same sampling methodology.

Nonsampling error

Nonsampling error may occur during the development or execution of a survey. There are several sources of nonsampling error. These errors can occur because of circumstances created by the interviewer, the respondent, the survey instrument, or the way the data are collected and processed. For example, errors could occur because:

the interviewer records the wrong answer, the respondent provides incorrect information, the respondent estimates the requested information, or an unclear survey question is misunderstood by the respondent (measurement error);
some individuals or businesses which should have been included in the census or survey were omitted (coverage error);
responses are not collected from all those in the sample (nonresponse error); and
forms may be lost, data may be incorrectly keyed, coded or recoded, etc. (processing error).

Information about nonsampling error for a specific program is provided or referenced with the data. The Census Bureau recommends that data users incorporate this information into their analyses, as nonsampling error could impact the conclusions drawn from the results.

Confidentiality

Title 13 United States Code, Section 9, prohibits the Census Bureau from publishing results in which an individual's or business' data can be identified. The Census Bureau's internal Disclosure Review Board sets the confidentiality rules for all data releases.

Title 13, United States Code: Title 13 of the United States Code authorizes the Census Bureau to conduct censuses and surveys. Section 9 of the same Title requires that any information collected from the public under the authority of Title 13 be maintained as confidential. Section 214 of Title 13 and Sections 3559 and 3571 of Title 18 of the United States Code provide for the imposition of penalties of up to five years in prison and up to $250,000 in fines for wrongful disclosure of confidential census information.

Disclosure Limitation: Disclosure limitation is the process for protecting the confidentiality of data. A disclosure of data occurs when someone can use published statistical information to identify either an individual or business that has provided information under a pledge of confidentiality. For data tabulations the Census Bureau uses disclosure limitation procedures to modify or remove the characteristics that put confidential information at risk for disclosure. Although it may appear that a table shows information about a specific individual or business, the Census Bureau has taken steps to disguise or suppress the original data while making sure the results are still useful. The techniques used by the Census Bureau to protect confidentiality in tabulations vary, depending on the type of data. For model-based estimates the Census Bureau uses other procedures to assure that the estimates and related information that are released cannot be used to disclose individual data.

Release of Source Data: Source data used in the production of model-based estimates are released to the public only when such release would not disclose individual data or violate other confidentiality restrictions applicable to the source data. This includes both source data obtained from Census Bureau surveys and censuses and also data obtained from other sources such as other government agencies.