ESTIMATING ???INDIVIDUAL DIFFERENCES??? IN THE RESPONSE TO AN EXPERIMENTAL TREATMENT