It is suggested that I use the intercoder verification in a research where I am the only coder, and whose objective is to verify the effect of the modification of the atmosphere of discussion on the topics discussed in a discussion group.
I believe it is useful to use test-retest to assess the reliability of my analyses (intracoder check), but I am not convinced of the relevance of using other coders to assess the codes assigned during my thematic analysis. If the objective of a research is to analyze the variations induced by the modification of a variable during almost similar discussion groups, using the same analyst, the same discussion plan and the same codebook, do you think it is necessary to check the agreement between coders?
ps.I work in a "codebook" or "template analysis" technique (as described in Braun and Clarke's (2022).
There really isn't much point in doing a test-test of your own coding because you will only demonstrate that you agree with yourself. As for inter-rater reliability, the value of that approach depends greatly on your analytic goals. In particular, if you will be counting codes as part of your comparison, then it is important to demonstrate that what you are counting is indeed reliable. Alternatively, if your are generating interpretive themes, then this is a subjective process where reliability of your coding is less relevant.
I don't think it's a must, as intercoder reliability is usually used when a team is involved in analysis. Even though coding involves a certain level of subjective judgement, intercoder reliability maintains that this judgement is shared amongst the members of the team.
One reason I can think of doing an intercoder realiability is to convince ( mainly the critics of qualitative research) the validity of the data related to content analysis. Also it is used if the researcher want the data to be coded in a consistent and standardized manner specially if it is required for a publication.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Artificial Intelligence