Non-Sampling Errors
Errors that are incurred from other sources than a selection of samples are known as non-sampling errors. Non-sampling errors may arise at the time of planning and execution of the survey and collection, processing, and analysis of the data. As well, non-sampling errors take place due to wrong selection of questions, wrong understanding and the wrong response of respondents, wrong method of research, and use of wrong tools for analysis of data. Non-sampling errors exist both in the census and sampling studies. Following errors are major non-sampling errors.
1. Errors of poor sampling design:
If the researcher fails to identify appropriate respondents and proper planning for selecting a sample then the results of the research cannot present the actual scenario.
2. Over and under coverage:
Sometimes, the researcher may select more elements while selecting a sample or may leave the essential elements of the population. Such a study cannot give appropriate results.
3. Misinterpretation of questions:
If the researcher uses difficult and ambiguous words, respondents may interpret the questions differently so that their responses do not represent the concept. Thus, a researcher should prepare a questionnaire considering the level of respondents otherwise there is a chance of a wrong response.
4. Processing errors:
There may be errors in coding, decoding, editing, and analyzing the data, and interpreting the results of the research. Such errors mislead the results of the research.
When respondents do not give a response or give bias response or not able to give a response or the researcher not able to record the response properly then the result of the research will be wrong.
6. Errors of researcher:
There might be errors in the research result due to weak definition of variables, selection of the wrong method, and preparation of weak questionnaire, and weak administration of questionnaire by the researcher.
7. Measuring errors:
Error many take place due to weakness in measuring. instruments like lack of awareness in respondents, unclear understanding of questionnaire due to poor preparation, unskilled, and untrained surveyor, such errors are measuring errors. Measuring errors also influence the results of the research adversely.
Methods of minimizing Non-Sampling Errors
Human errors are obvious in every human activity but to make research work more reliable and valid, such errors should be minimized. It is very hard to minimize the non-sampling errors. The researcher can minimize such errors doing following works:
1. Checking the sampling process:
Check the elements of the population while selecting for research. It is down so as to find out whether the essential elements of the population are left or unnecessary elements are included or elements are repeated. Such a check can be made by the researcher himself or with the help of experts.
2. Preparation of questionnaire:
The researcher should prepare a questionnaire considering the level of respondents. The questionnaire should not contain dual meaning, unclear and ambiguous words, and complex sentences. The questionnaire should be clearly and easily understandable to the respondents.
3. Pilot survey:
Conducting a pilot survey helps to reduce the problems related to the questionnaire. It helps to get feedback and improve the questionnaire.
4. Fix procedures:
Establishing certain procedures while involving respondents and the involvement of researchers in the research work helps to reduce non-sampling errors.
5. Use of competent manpower:
A competent surveyor or interviewer is to be selected so that they can collect appropriate information or data.
6. Provide information:
The surveyor should tell his/ her experience or related facts or other informal issues to the respondents to draw their attention but should not talk about the weaknesses of the researchers.
7. Provide training:
Providing training to surveyor related to the filling up of questionnaire, method of taking interview, behavior to be presented with the society and individual and method of concluding the interview or other works.
8. Use of experts:
The researcher should take the help of an expert to minimize errors, especially in coding, recording, and decoding.
9. Checking data processing and analysis:
The researcher should use various techniques to check all the steps in the processing and analysis of data. Such a check helps to strengthen the work of data processing and analysis that helps to reduce errors.
