Michael Sosin's "Delivering Services under Permanency Planning" (Social Service Review 61 [June 1987]: 272-90) contains interesting data about the delivery of services in Wisconsin under permanency planning. Unfortunately, Sosin commits a classic academician's error-conducting a careful analysis on a dependent variable that is so narrowly defined as to mislead.1 Whereas the concept of "permanency planning" lacks consensual meaning, I think few would agree with Sosin that it can be boiled down to "the speed at which children are discharged from care" (p. 275). Children deserve to have permanent plans made as soon as possible, but some children will be in care a long time for good reason. Adoptions should not take years to complete, but they often do. A "fast adoption" is virtually an oxymoron, yet the great majority of adoptions successfully promote children's long-term developmental needs and result in permanence.2 Wisconsin should be praised for being among the nation's leaders in preventing the need for long-term foster care through the use of adoption, yet Sosin discounts efforts to terminate parental rights and make adoptions possible as "spending time on low-probability outcomes" (p. 285).3 Sosin's study has other significant shortcomings as does most child welfare research.4 None is so fundamental as the confusion between speed and accuracy as the goal of permanency planning. Sosin's suggestion that workers should engage in less discussion of parental-right terminations and adoption plans with birth parents because they are "inconsistent with quick discharges" (p. 285) is the result of such confusion. Birth families must hear that the failure to complete the reunification plan is grounds for termination of their parental rights. Also, Sosin's suggestion of a causal relationship between such discussions