Do we need validity? –A critical appraisal of validity in qualitative research.

Abstract

Background: Validity is understood to be a question of the correctness or accuracy of research. It can be viewed as a condition for science’s aspiration to produce authoritative knowledge. When it comes to qualitative research, the claims produced are often interpretations of people’s way of understanding. The validity of qualitative research can therefore be viewed as particularly difficult to affirm.
Purpose: The purpose of this thesis is to review literature on validity in qualitative research. It also discusses the literature in relation to using it in assessment of research. Literature was collected through extensive searches in method books on qualitative research and databases. The focus is on research in sociology.
Results: Validity in qualitative research has been understood, among other things, as truth, accurate representation and adequacy. It has also been reconceptualized as concerning success in practice and acknowledgement of multiple possible interpretations. An attractive solution to the problem of validity is to equate it with that findings can be trusted to use in further research. A common theme is that validity is dependent on consensus, e.g., consensus concerning the meaning and application of concepts. Problems with consensus is that it can be hard to reach, standards based on it may be hard to practice, and it can also stop innovation.
Conclusion: A workable and mutual concept of validity is possible if there is consensus within a research community on, e.g., concepts and their application. However, such a consensus may block new interpretations and work as a “policing” of a research field. Instead, conceptions of validity should be treated as incitements to research. Moreover, a generic conception of validity, independent of what purpose the research should serve, is not possible. Still, transparency and honesty is generally called for, since they facilitate the evaluation of research.

Popular Abstract

That a research project has validity has generally been understood as that its findings are true or correct. If a research study has validity, then what it finds is correct. What it says is how things really are. Qualitative research in the social sciences studies the social world with the help of, among other things, interviews, observations of people in their everyday situations, and analysis of texts. This thesis is a review of what validity is in qualitative research. It reviews what others have written on the subject. It also discusses these conceptions of validity in relation to how they would work as guidelines for doing research, or for someone who is reading research and wants to evaluate it.
What this thesis finds is that there are numerous descriptions of what validity is. A great deal of descriptions say that research has validity if it describes and explains something as it really is. Others say that the interpretations of qualitative research cannot be true. They can only be “adequate”, in the sense of agreeing with how we typically understand something. Still others have likened validity to trying out the practical meaning of knowledge, for example, to see if it can be used to be successful in a practice. Validity has also been understood as being dependent on what the purpose of the research is. For example, research that has the purpose to help people to better understand their lives is valid if it does this. There have also been reactions against validity as correct interpretation. Some writers hold that an interpretation is always incomplete and that other interpretations are always possible. Later texts on validity say that validity is about the quality of the whole research process, and not just the quality of the final research claims.
This thesis concludes that it is possible to have a workable general idea of validity if there is agreement. Researchers that agree upon how to describe experience can form standards based on this. They can use these standards to decide the validity of research. However, such an agreement could be hard to reach. An agreement on standards could also mean that research that is original and useful but breaks these standards is rejected. A more useful view is that conceptions of validity can serve as optional guidelines for researchers. They can as such be of help to researchers in their work. Further, researchers should always be honest and show how they have done their research. Readers of research reports can then see and judge the steps that led to the researcher’s interpretations.

• author: Kihlgren, Per
• supervisor: Veronika Burcar Alm
• organization: Sociology
• course: SOCM04 20161
• year: 2016
• type: H2 - Master's Degree (Two Years)
• subject: Social Sciences
• keywords: validity, validation, qualitative research, method
• language: English
• id: 8886747
• date added to LUP: 2016-08-24 12:21:44
• date last changed: 2016-08-24 12:21:44