|
Chapter 9.
Occupational Safety and Health Statistics
Reliability
of Estimates
All estimates derived from a sample survey are subject to
sampling and nonsampling errors. Sampling errors occur
because observations are made on a sample, not on the
entire population. Estimates based on the different
possible samples of the same size and sample design could
differ. The relative standard errors, which are a measure
of the sampling error in the estimates, are calculated as
part of the survey's estimation process. Both the
estimates and the relative standard errors of the
estimates (or statistical models for approximating those
relating to worker and case characteristics) are
published in appendix A to the annual BLS bulletin, Occupational Injuries and Illnesses:
Counts, Rates, and Characteristics.
Nonsampling errors in the estimates can be attributed
to many sources, e.g., inability to obtain information
about all cases in the sample, mistakes in recording or
coding the data, definitional difficulties, and so forth.
Although not measured, nonsampling errors will always
occur when statistics are gathered. To minimize the
nonsampling errors in the estimates, the completed forms
are edited and apparent inconsistencies are checked with
the employer, who is encouraged to respond fully and
accurately to all survey elements. A small fraction of
the sample does not submit usable data. To account for
this missing information, nonresponse adjustment factors
are applied at the appropriate industry and size class
level.
Next: Publication
Guidelines
|
Bureau of Labor Statistics
