Social science has been defined as ‘the attempt to explain social phenomena within the limits of available evidence’ (Lewins 1992: 41). The definition of what counts as ‘evidence’ is problematic, as is the question of how evidence is collected. Evidence has traditionally been understood as data collected through the counting of responses or observations. In circumstances where sociological concepts were not directly observable, they were instead grasped indirectly through the testing of hypotheses where ideas could be viewed as ‘facts’ if they were verifiable by certain observations. Society was seen as rule governed, with there being one ‘true’ reality, and research was based upon the idea of there being one universal knowledge which was created through the application of objective thinking and empirical research. Emphasis was placed on a reasoned, scientific approach to knowledge production, wherein any researcher operating under the same conditions would reach the same conclusion (Code 1993). However, attitudes about what constitutes valid and reliable knowledge, research and evidence have changed as social research has developed.