fix bug in roc_solver.m that incorrectly output the dprime1_lure para…

…meter of the MSD model. This parameter was set to the most conservative response bias parameter. This had not impact on the underlying model parameters, so if the MSD model was estimated without the dprim1_lure parameter varying, there is no impact on the resulting parameters.

main/roc_solver.m

@@ -511,7 +511,7 @@
     data.(modelField)(index).parameters.Dprime1_targ = ...
         bf_pars(:,2) + bf_pars(:,4);
     data.(modelField)(index).parameters.Dprime1_lure = ...
-        bf_pars(:,7) + bf_pars(:,9);
+        bf_pars(:,7);
 end
 
 % Generate the predicted and fitted data