Table 2.
Sampling and Nonsampling Errors Associated with Surveys and Their Causes1
| Type of Error | Definition | Causes |
|---|---|---|
| Sampling error | Error that arises when the survey sample is not representative of the population to be evaluated | Inappropriate choice of sampling frame |
|
| ||
| Nonsampling errors | ||
| Measurement error | Difference between the response recorded by the individual and the true value | Unclear questions; inadequate training of interviewers; inappropriate mode of survey administration; inaccurate translation (for surveys administered in multiple languages) |
| Processing error | Inaccurate coding of survey responses | Lack of a data dictionary; inadequate training of coder(s); use of multiple coders |
| Nonresponse error | Error caused by missing responses on some of the questions in the questionnaire (partial nonresponse) or the entire questionnaire (total nonresponse) | Questions about potentially sensitive topics; inappropriate mode of survey administration for the selected sample; poor design of questionnaire; response burden (questionnaire too long and/or too complex) |