Transfer of problem-solving strategies between dissimilar problems, or problems of different content,is notoriously difficult to achieve even for adults (Gick &Holyoak, 1980;Hayes & Simon,1977; Tuma & Reif, 1980).

It is this generalized problem-solving transfer that computer programming educators hope to increase. Certainly, that is one of the claims made about the benefits of Logo instruction (Papert, 1980). Ginther and Williamson (1985), however, state that this type of generalized problem-solving transfer is difficult, if not impossible, to achieve in any context, and there is no reason to claim that programming language instruction will provide contrary evidence.