There are various kinds of errors that have a greater chance of occurrence in Questionnaire based research. These four kinds of errors are:
a) Sampling Error
b) Measurement Error
c) Coverage Error
d) Non- Response Error
a) Sampling Error: This kind of error largely deals with Inadequate Sample size and non-random Sampling techniques used.
b) Measurement Error: This kind of error occurs at the stage of designing questionnaire. If the questionnaire is imperfect or faulty, it comes in the category of Measurement Error.
c) Coverage Error: This error occurs at the data collection stage of questionnaire based research. It talks about inability to due to multiple factors in being able to contact some people in the sample universe.
d) Non Response Error: This kind of error arises when people of a particular familiar trait are not covered in the sample size because they have a common non responsive trait.
With the scope for all these errors, the researcher needs to construct effective validity. There are different types of validity that needs to be considered by a researcher. They are:
- External Validity: This type of validity assists in examining if an observed causal relationship can be given a generalised conclusion across multiple measures, persons, settings and times. It specifically talks about either generalising to a well specified population or across various sub populations.
- Statistical Conclusion Validity: It is focussed upon the power to be able to identify the different relationships and the magnitude of relationships. One of the major hurdle or setback at this stage is to have an in adequate sample size. When this happens, the sampling error goes very high and hence the validity of the conclusions drawn from the study are weak and so are the inferences.
Certain Compliance principles should be kept in mind while designing these survey studies. They are:
a) Reciprocal perspective: Respondents usually have the willingness to comply with a request provided the request to participate in a survey has a reward system associated with it in terms of a gift, favour or concession.
b) Consistent approach: After having given the approval for the position, respondents are often found willing to co-operate and comply with the requests for behaviour that have consistency with the position.
Social Validation: Often people are found using beliefs, attitudes and actions which have a certain trait if similarity for their personal beliefs, attitudes and actions. They are often found showing willingness to participate as respondents if assured that people falling in the similarity bracket would be doing the same.