I applied this to 3 studies from Piray’s review (selected only due to availability of empirical Dirichlet params in public materials). Power remained high under substantial attenuation, only dropped to the low levels estimated by Piray under 90% attenuation of the empirical population preferences