Hi @tashrifbillah,
I think the reconstructed reference data should be added to the statistical comparisons in the box plot comparisons and csv files. This will be very beneficial because the users/we are generally using the reconstructed reference data instead of the original reference dwi.
We can still keep the original reference to see how much spherical harmonics fitting affected the FA, but adding reconstructed will improve the comparisons.
Thanks!