Destruction of the Müller-Lyer Illusion as a Function of Procedure and Pretraining

Abstract

Two experiments were performed to investigate conditions affecting the Müller-Lyer illusion and its decrement with practice. The first experiment was a methodological study concerned with the setting of the variables before adjustment by the subject. The results indicated that the method employed may determine whether a decrement occurs with repeated trials. The evidence suggested the most suitable method to employ in the succeeding experiment. The second experiment was performed to investigate the effects of practice with another illusion figure on the magnitude of illusion on the Müller-Lyer figure. The practice figure was the same as the Müller-Lyer illusion figure except that circles were replaced by the obliques. It was found that the magnitude of the initial illusion is a decreasing monotonic function of the amount of preliminary training. This finding is interpreted as meaning that in pretraining subjects are practiced in disregarding the context (circles) of the horizontal lines of the figure and this transfers positively to the Müller-Lyer task.