Datensatz als E-Mail versenden: Estimating the Causal Effect of Randomization Versus Treatment Preference in a Doubly Randomized Preference Trial