“…Donaldson (1978), for instance, found that although three to four year old children may face considerable difficulties with the procedures of a problem task they are given in a particular context that is unfamiliar to them, and that when the same problem is presented in a familiar and realistic setting this results in dramatic improvement in performance. On the other hand, numerous studies have shown that deliberately setting problems in familiar contexts has resulted in children reverting to familiar everyday systems of explanation, abandoning learnt scientific knowledge (e.g., Solomon, 1983;Bliss, Morrison, & Ogborn, 1988). In the latter studies learners made greater use of scientific principles if questions were set in obviously scientific contexts, demonstrating that the smaller the distance between the context in which learning takes place and the one in which new knowledge is to be used, the more likely it is that such transition will take place successfully (Engel-Clough & Driver, 1986;Murphy & Schofield, 1984;Toh & Woolnough, 1994).…”