I am comparing two devices by Bland-Altman plot. However, when I am looking other papers, there are also using: Pearson correlation, ICC, standard error of measurement, cronbach's alpha, ... Is it useful to use other methods than Bland-Altman? If yes, which one should I use to improve my comparison study?
No, only ICC can help. Others are irrelevant. Take a look at Bland and Altman's original papers and the following links
http://www.jerrydallal.com/lhsp/compare.htm
https://www-users.york.ac.uk/~mb55/meas/meas.htm
Dear Yuann,
You may use Pearson correlation if you are interested to find "linear" relation between your variables. If this is the case, you may apply it to check.
Standard error of measurement is essential for sure. It shows reliability of your test. However, if you use ICC, you may skip it. Because ICC itself shows you the reliability of your measurement.
Since you are studying only 2 specific groups, using cronbach's alpha is not related to your work.
Good luck
How to calculate ICC? I am looking online and McGraw (1996) gives 5 ways while comparison studies do not mention the way to calculate it...
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics