About a qualitative content analysis, many scholars suggested intercoders reliability as a measure to evaluate reliability (very common). A question is, what if the whole transcripts say 20 respondents were coded by only one coder? In your opinion, what are the means to establish a reliability (stability) for a single coder?(e.g., intra-reliability test). Any suggested reading materials or articles?
Hi,
With 1 coder, there can't be an inter coder reliability therefore it means that your results may not be reliable and subject to bias.
To achieve reliability in your analysis, get another coder, preferably one with no prior knowledge of the research aims and objectives to code. The main coder could be you/the author but in best practice, it's best to get another person, even for just the 20 analysis.
Hope this helps.
Hi emmanuel, thank you. I said intracoder, not inter though. Many literature on intercoder, but as far as i know, even one coder, in qualitative, reliability or dependability can still be achieved.
There is a reason why people emphasize inter-coder reliability, because it is all about a second coder's ability to reproduce the original coding. Independently reproducing the same results is the essential definition of reliability.
Yes, one coder can become more consistent over time, but if another coder cannot produce the same results, this is mostly a matter of getting better at perceiving what you set out to perceive. For example, if you finished your coding and then went back and re-coded the first interview, that would just speak to the evolution of your own coding technique, rather than anything that would meet the technical definition of reliability.
Of course, there is always the question of why you would want to establish reliability, which is basically a quantitative concept.
Prof David, Thank you for your advice. You are right where reliability is meant for quantitative concept. I saw people or some papers in qualitative research, they established intracoder reliability like what you mentioned "if you finished your coding and then went back and re-coded the first interview", Get a percentage. So, i would like to ask, can i employ that technique since you said it is more towards evolution of coding technique. Or perhaps other measures in establishing "reliability" in qualitative research. TQ
I think you can establish the reliability of your qualitative research for example interview, if your were speaking through the voice of the respondents and not only your own voice. That will help to establish the comprehensiveness, coherency and trustworthy of the research findings, and that is what reliability seek to confirm in a study
Hi, intra-coder reliability can be assessed by using 2-way mixed intraclass correlation coefficients.
Coding at two different periods of time to establish reliability is test-retest method of reliability.
- Badges
- Science topic
- Similar topics
- Social Psychology
- Attitude