I am measuring immunofluorescence intensity using image analysis techniques. I did this twice and would like to find out how different/consistent my two data sets are?
What I described was the Standard Error of Measurement (SEm), a measure of repeatability. You can calculate repeatability standard deviation without any special tools, or have a look then at the BlandAltman function in the R package 'ResearchMethods'.
I understand you have several immunofluorescence samples and you did two measurements on each under the same conditions.
As a first step you could visualise your data. Put on the horizontal axis the mean of the two measurements, and on the vertical axis the two intensity measures themselves. This will give an idea if the level of consistency depends on the intensity or not; if not what transformation (often log) makes the level of scatter independent from the mean.
After the transformation (if one is necessary to achieve homoscedasticity) you can run a one-way ANOVA, where the the sample id is the blocking factor. This will tell you the within sample standard deviation, which is a useful measure of the intra-observer retest variability.
What I described was the Standard Error of Measurement (SEm), a measure of repeatability. You can calculate repeatability standard deviation without any special tools, or have a look then at the BlandAltman function in the R package 'ResearchMethods'.
I've been looking and, the ICC test use a similar formula. Indeed, repeatability test is also called "interclass correlation coeficient" and that's the ICC, is the package "irr" or "psych".