We report on fits of the SU(3) N_f=8 LSD spectral data to chiral
perturbation theory with a dilatonic meson. These fits confirm that
current simulations are in the ``large-mass'' regime, with approximate
hyperscaling as the leading mass dependence. We find that the
leading-order effective field theory describes the data well. In
particular, the effective field theory allows us to understand the
staggered taste splitting, explaining the pattern observed in the LSD
data, which looks different from QCD.