prev next front |1 |2 |3 |4 |5 |6 |7 |8 |9 |10 |11 |12 |13 |14 |15 |16 |17 |18 |19 |20 |21 |22 |23 |24 |25 |26 |27 |28 |29 |review
- With regard to health promotion, McQueen and Anderson argue that in fact "evidence" may be viewed as an "inappropriate" idea. In support of this view, they suggest that the underlying epistemological base for health promotion is not experimental science but rather the social and behavioral sciences which by and large, are non-experimental, highly empirical, and characterized by observation and classification, rather than by any effort to manipulate the subject of study. Furthermore in those situations where variables are being purposely manipulated, the design can rarely possess the rigorous structure of an experiment because of the lack of control over the relevant variables and the inability to have a control group for comparison. The most we can usually hope for is a quasi-experimental design

- To make things even worse, it has been argued by the WHO Working Group that a key characteristic of health promotion evaluations is that they involve "participation" , ideally on an equitable basis, of both researchers and the researched (http://www.who.dk). According to McQueen and Anderson,this characterization of health promotion and the attendant assumptions would make a rigorous experimental design totally inappropriate and, because the term "evidence" is so closely associated, both historically and in the conceptual mentality of researchers, with rigorous experimental designs, the use of the term in health promotion becomes highly questionable. They further argue that the casual use of the term "evidence" in health promotion presents a challenge because in their view, it will seriously deflect health promotion practice from concentrating on how best to design and evaluate interventions

prev next front |1 |2 |3 |4 |5 |6 |7 |8 |9 |10 |11 |12 |13 |14 |15 |16 |17 |18 |19 |20 |21 |22 |23 |24 |25 |26 |27 |28 |29 |review