In each of these studies subjects were called upon to select the most preferable object from several identical objects. The object in each case was a product, such as bread or men's suits, which was beyond the power of these individuals to evaluate completely from direct observation and physical contact alone. It seems reasonable to expect that, under these circumstances, individuals will infer unobservable characteristics of the focal object from the observed behavior of others with respect to that object. That is, it seems likely that the observed behavior of others with respect to the object will be informative about that object.