On Task Context and Set Solutions to Water-Jar Problems

Abstract

A test of 6 training and 20 main problems elaborated on a model similar to the model which underlies water-jar problems is described. The training problems and instructions require that each problem may have more than one solution. The 20 main problems can be solved by the same formula (set solution), but 10 of these problems (the test problems) can also be solved by a shorter formula (non-set solution). The score is the number of non-set solutions for the 10 test problems. A comparison between an individual condition of testing for 21 subjects and a collective condition (40 subjects tested in a group) shows a significant difference in frequency distributions of scores. More subjects gave a small number of non-set solutions in the group condition than in the individual condition. For the two conditions, data show significantly more non-set solutions to the last test problems than to the first test problems.