A recent piece of scholarship concerned with film audience reactions suggested that the pleasures derived from cinemagoing are not homogeneous. At a general level, De Vany and Walls presented evidence to support their contention that 'Film audiences make hits or flops and they do it, not by revealing preferences they already have, but by discovering what they like'. The implication is that, as hard as film-makers attempt to attenuate the risks associated with film production through the devices of 'stars', 'genre' and perhaps 'direction', audiences are prone to upset the best-made plans. However, this can be taken further with Harper and Porter's claim that, for the period of their study (circa 1950), '... gender, social class and, to a lesser extent, age were paramount in determining the nature and intensity of film response'. They argued that, although films and their stars are important in shaping the experience of audiences, the latter, in turn, bring their own experiences to the cinema and these act as filters through which meaning and pleasure are derived. From their analysis of the 1950 Mass Observation study into the incidence and reasons behind crying at the cinema, Harper and Porter found that filmgoers, classified into gender and social class categories for the purposes of analysis, gave differentiated responses to both the act of crying and the films which stimulated this behaviour.