The concept of reliability in qualitative research needs to be studied keeping in mind the interpretivist or naturalist paradigm of the research. In positivism or quantitative research, the use of reliability and validity is very common but the questions always arise that how to maintain the concept of reliability in qualitative research studies. There are various ways in which the reliability can be enhanced in a qualitative research. Before indulging into the details about the concept of reliability in qualitative research let’s discuss some basics about the reliability in the research.
Definition of Reliability in research
According to Moser and Kalton, “a scale or test is reliable to an extent that repeat measurements made by it under constant conditions will give the same result.” and the reliability of a research instrument or tool is defined by Kumar in following words, “if a research tool is consistent and stable, and hence, predictable and accurate, it is said to be reliable.”
How to measure the reliability of research instrument
To measure the reliability of a research instrument the researcher has to collect the same set of information from the same audience under similar circumstances more than once if the result yielded has great comparability it is said that the research instrument is very reliable, on the other hand, if the results are not comparable it is said that the research instrument is hardly comparable and not reliable. Therefore, two things are important in considering the reliability of a research instrument:
- How reliable the research instrument is;
- How unreliable is the research instrument?
Concept of reliability in qualitative research
- In qualitative research, repeatability is often not the aim of the research and when the research instrument such as a questionnaire is used repeatedly over a period of time it is difficult to get the same responses. Several factors influence the respondents to respond differently. A simple reason can be their experiences which usually change over a period of time with new experiences respondents may start thinking differently. The researcher has no control over the experiences of the respondents there are many other factors that he/she cannot control.
For example, you conduct interviews to know the prevalence of juvenile crimes in a community. You ask questions from some people that you selected randomly and you came to the conclusion that 5 % of the respondents say that they were physically or emotionally harmed by the adolescents. Now you decide to conduct the same interview again after a lapse of say a month and you get to know that 10 % of the respondents say that they encountered or became a culprit of a juvenile crime. The reliability of this research instrument is not good. There can be various reasons for the unreliable results in this case; one of the reason could be that new criminal acts happened in that community. To make a research instrument reliable the researcher can take several re-tests before reaching to any conclusion.
How to enhance the reliability in a qualitative research
There are various ways in which the reliability of the research instrument can be enhanced although it is impossible to ascertain that you get a 100 percent reliable research instrument in a qualitative research.
- The place where the interview is done or the questionnaire is submitted or in the case of observations where the observations are made should remain same in re-test.
- The respondents can also affect the reliability of the research instrument if they are in a different mood in different interviews. Making the environment as neutral as possible can bring positive effects.
- The way the interviewer interacts with the respondents should remain same in each interview.
- The wordings of the questions themselves can make the research instrument reliable or unreliable. Simple a reliable question is one that is simply stated and does not have double meanings. This means that different respondents should not take different meanings of the questions that you have phrased for them.
- There can be other ways in which the reliability can be enhanced for example the researcher can take the test several times before reaching any conclusion.