The difficulties I continue to have with grounded theory include the fact that phrases like "researchers enjoy what can be done with data once they are collected' (Strauss and Corbin 1998: 5) imply a separation between collection and analysis, whereas later the authors insist that concepts and design must be allowed to emerge from the data', implying the iterative approach I advocate in this book. Indeed I would describe their approach as iterative and would suggest that the emphasis on induction belies their actual methodological insistence on being both deductive and inductive. Ezzy (2002) reminds us, however, that qualitative researchers' early insistence that the work is inductive was a response to the hege mony of the deductive approach, and that most research actually draws on both inductive and deductive reasoning A further problem that is not well addressed is how theories become linked together in this approach. There is an implication that every piece of research generates new theory, which does not seem very helpful on a general level.However, there is some small acknowledgement that some research might actually be to test theory.