Hypothesis Salience, Task Difficulty, and Sequential Effects on Problem Solving

Experiment I required four groups of subjects to each solve three closely related problems. The problems differed in difficulty. Groups RSC and RCS were given problems in order of increasing and decreasing complexity respectively. All problems could be solved by the application of the same simple rule. Speed of problem solution was facilitated by giving subjects the simple problems first. Group RSC solved their problems more rapidly than Group RCS. This is referred to as a Groups TSC and TCS were also given their problems in order of increasing and decreasing complexity. The problems in this case were of a nature that induced subjects to solve them by a trial and error process rather than by rule induction. No sequence effect was obtained in this case. Experiment II indicated that effectively insoluble problems could be solved very rapidly if preceded by the appropriate simple problems when using rule-based tasks, giving a massive sequence effect. This could not be replicated in Experiment III using problems solved by a trial-and-error process. It was concluded that the sequence effect could only be obtained on rule induction problems and that its occurrence could be predicted from Hypothesis Testing theory. It was further concluded that these results provide additional evidence that the sequence effect and Einstellung are complementary phenomena.