The effects of subject arm position and initial experience on Chevreul pendulum responses

Some clinicians maintain that responses to the Chevreul pendulum illusion are facilitated by resting one's elbow on a table. Others claim the reverse. We compared these two methods in a counterbalanced crossover design by having 32 university students perform the Chevreul pendulum illusion with their elbows supported on a table and with their elbows unsupported. Although there was no main effect for method (elbow supported versus elbow unsupported), subjects who rested their elbows on a table on the first trial were more successful in responding on both trials. This suggests that supporting the elbow does facilitate responding, but only on the initial trial. Performance on subsequent trials is determined by degree of success on the first trial. Similar data from a previous study comparing different hypnotic inductions suggests that this phenomenon is generalizable beyond the Chevreul pendulum illusion and supports the hypothesis that the test-retest reliability of suggestibility scales may be due to a stabilization of response expectancy by a person's first experience of imaginative suggestions.