Definition: The Non-Sampling Error is the statistical error that arises due to the factors other than the ones that occur when the inference is drawn from the sample.
Simply, the errors caused due to the defective methods of data collection, faulty definition, incomplete population coverage, wrong tabulations, etc. when the complete enumeration of all the items in the universe is surveyed is called the non-sampling error.
Often, investigator expects that the research will be free from errors when all the items in the population are studied, but practically, this is not possible. As it is very difficult to avoid the errors of observations and ascertainment and also the tabulation errors while processing the data can invalidate the results obtained.
Due to the complete enumeration survey, the non-sampling errors are likely to be more than the ones arising out of the sample survey. But, however, the non-sampling errors can be reduced to a great extent if organized and trained personnel are used at the field and tabulation stages.
With an increase in the sample size, the non-sampling errors are likely to increase, which is opposite to the behavior of a sampling error that reduces with the increase in the sample size.
Some of the major reasons that lead to the non-sampling errors are:
- Inadequate data specification or data being inconsistent with the objective of survey or census.
- Inadequate methods of data collection.
- Duplication of a subject in the survey.
- Lack of trained investigators.
- Lack of supervision of primary staff.
- Errors committed while tabulating the data.
- Inadequate scrutiny of primary data.
- Errors due to non-response of the subject.
- Errors in data processing Viz. Coding, punching, verification, tabulation, etc.
Note: The results obtained from the complete enumeration of data, although free from the sampling error is subject to non-sampling error, while the results obtained through a sample survey is prone to both the sampling error and the non-sampling error.
Leave a Reply