Bioequivalence and Bioavailability Forum

Dear all,

thanks to Jiři Hofmann bringing log₁₀-transformation (which is not uncommon in SE-Asia and Australia) back to my attention.
Example data: Table 3.6.1 of Chow/Liu, MSE

ln-transformed:    0.037221
log10-transformed: 0.0070202

CV_intra = 100√ℯ^MSE–1 = 19.47%
Now for the gem – in the case of log₁₀-transformed data we cannot simply back-transform like
CV_intra = 100√10^MSE–1 = 12.77%
Instead we have to apply:
CV_intra = 100√10^ln(10)×MSE–1 = 19.47%

---

Based on a discussion with Linda Hughes / Pharsight.