Two parameters measure apparent change during repetitive stimulation

In two studies subjects observed a 127 mm skeletal white diamond rotating around its vertical axis while binocularly fixating a white dot immediately behind it and 1.273 m away against a black background for 10 min. They spontaneously reported many different apparent changes. Contrary to previous studies, reversals of direction of rotation were neither the first nor the most common change and sometimes never occurred. When reversals were not stressed and reports of any change were requested, the latency of the first response correlated with the total number of new form reports but neither correlated with the transition rate between forms. On repeated testing all three measures are highly reliable. The rate at which new forms appear is therefore independent of transition rate. Current ways of describing similar effects by smooth curves fitted to group data confound the two parameters, or fail to recognise that nonreversal reports are very common.