What is the statistical test used for checking inter observer agreement in case of a continuous variable (Eg. Measurement of height)?
Cronbach alpha might also work for your case.
cheers,
Marco
Depending on the characteristics of your variablem, you could use a many-facets Rasch model:
http://ltj.sagepub.com/content/15/2/263.short
http://www.jstor.org/stable/1435170
If this not apply to your situation, you could check intraclass correlation as Rimantas Vosylis sugest
Hi, Vishnu
ICC requires normality of data set distribution. Moreover, it shows a paradoxical inflation when is used on heterogeneous samples, with a wide range of values. Both issues are often overlooked, and the second effect is a real limitation to assess reliability. That´s the reason why it should not be employed as the sole statistic.
In my opinion, the best choice is the use of ICC and the Bland-Altman method: Whereas ICC is a global measurement of reliability, the Bland-Altman method provides an interesting information about the distribution of differences.
Cheers,
Emilio
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology
- Biosignals
- Biomedical Signal Processing
- Statistical Analysis