I am designing a study where I measure the electrical conductivity of tissue in two directions. My (pseudo) hypothesis is that, regardless of the magnitude of the measured and predicted conductivities, the anisotropy ratio should be the same for model and measurement.

Each measurement and model prediction should be approximately normally distributed. So I am trying to determine if my model of anisotropy ratio matches the measurement.

I am leaning toward Bland-Altman analysis because I view these as two different methods of arriving at the same value. If I am going to use something like a t-test, I realize that I will have to do the paired version because each measurement location is likely to be different.

Am I on the right track with Bland-Altman? Is there anything I am missing about ratios of supposedly normally distributed values? Should I be using a propagation of uncertainty approach to determine the variances of the data sets?

Any help you could give me would be much appreciated.

Noel Montgomery