There is a more detailed post on this project here: https://www.researchgate.net/post/What_are_the_proper_techniques_for_analyzing_reliability_of_categorical_variables_with_a_large_number_of_categories, but I wanted post a new, more specific question in a new thread.
For those of you used to looking at I&O code reliabilities (or using I&O codes in substantive research), what levels of agreement/reliability are acceptable to you? Or, what levels have you commonly seen? I'm using Census, NAICS, and SOC codes, and seeing ARs and kappas between mid-high 0.70s and mid-high 0.80s. Do those seem high, low, or average? This is my first time working with I&O data.
Thanks in advance!
I think inter and intra kappa should work but then you can download this paper for more information
Inter-rater agreement of assessed prenatal maternal occupational exposures to lead.
Correa A1, Min YI, Stewart PA, Lees PS, Breysse P, Dosemeci M, Jackson LW.
- Badges
- Science topic
- Similar topics
- Econ
- Manufacturing
- Industry