Non-sampling error

Non-sampling error is a statistical error that arises in the data collection process and causes the data to differ from the true values. It is not related to the sampling process and can occur even in a census. Non-sampling errors can be caused by a variety of factors, including:

1. Response errors: These occur when respondents provide inaccurate, incomplete or biased answers. For example, a respondent may not remember information correctly or may choose to provide false information.

2. Measurement errors: These occur when the method of obtaining measurements is flawed. For example, if a survey question is poorly worded, respondents may not understand what is being asked and provide incorrect answers.

3. Processing errors: These occur during the preparation and management of data. For example, data may be incorrectly entered into a database or mistakes may be made during the coding process.

4. Coverage errors: These occur when some members of the population are not considered in the survey. For example, if a survey is conducted online, those without internet access will not be included.

5. Non-response errors: These occur when individuals chosen for the sample do not respond or cannot be contacted. The characteristics of these non-respondents may differ from those who do respond, leading to bias in the results.

To minimize non-sampling errors, it's important to carefully design and pretest questionnaires, train interviewers, and carry out quality checks.