Non Sampling Error In Statistics

Sampling error
It is the difference between the selected sample from the true representation of the entire population that is the true characteristics, traits, behaviors, qualities or figures. It always occurs when the researcher draw different conclusions from the sample population which is a subset of the entire population and it has its individual different in them. Most frequent cause is the biased sampling procedure in which a sample is not true picture of the entire population. Another possible cause is chance, which is the process of randomization and probability sampling. This leads to systematic error.
Non-Sampling error
It is an error that relates to all other types of survey errors other than sampling. These errors come from the

Questionnaire error
This type of research error may occur in many forms:
Content of the Questionnaire
A questionnaire error may occur when the wording in the questionnaire is either wrong or one uses difficult terms that may be misunderstood by the respondent. It may also occur due to typo errors.
The ordering and format of the questionnaire may also lack a proper arrangement or flow. The respondent may be confused, for example, one may not know whether the questions are related or each stands on its own.
Method of delivery
The method of delivery may sometimes lead to questionnaire error. For example, when using a telephone, some words from both ends may be heard incorrectly and leading to wrong collection of data.
Inappropriate and vague measurements may also be used in the questionnaire that may be interpreted differently by different respondents. For example, ‘How would you describe your salary: high or low?’ High or Low is relative and therefore the measurements are vague and faulty.
Questionnaire structure error
This occurs when the flow of topics, the sequence of the questions as well as the formulation and wording in the questionnaire have not been used properly. This means that the questionnaire will not be understood by the

Example, the interviewer does not ask the sensitive questions related to a respondent but later fills in the answers based on personal assessment. It also occurs when an interviewer falsifies entire questionnaire or fills in answers to questions that respondents skipped.
Some interviewers cheat to finish an interview as quickly as possible or to avoid questions about sensitive questions.
Under cheating error we have curb-stoning which refers to interviewers filling in responses for respondents that do not really exist.
Research firms can reduce cheating error by contacting a small number of respondents to verify previously recorded responses. In this way the curb stoners might get revealed and face the consequences associated with this unethical behavior.
Population definition error
Is a type of research error where the researcher wrongly defines or establishes the population being studied hence wrong conclusions. For example the researcher may be trying to study the behavior of teens in using the internet but then defines teens as any one above 15years but less than 28 years. However, according to the law teens are those up to