Problem solving in sciences and mathematics is usually taught
by transfer strategies: a set with a few problems is fully explained and then
the students are expected to solve analogous problems. Secondary teachers
frequently assume that building the analogical relationship among solved
problems and the target ones are an easy cognitive task. So they claim that
the cause of the low level of success in science problem solving is the lack
of mathematical procedural competence. In this paper we perform an
experiment to probe whether this claim is the main cause for the failure in
science and mathematics problem solving or not. The results show the main
cause will come from the failure in the construction of the appropriate
situational-model and/or the problem-model representations.