Do ambiguous images provide psychological insights? Testing a popular claim