Masters Dissertation Example

developing country studies. The majority of resident co authors had NGO/Govt affiliations, but this did 
not result in higher levels of implementation 
To highlight this, the few studies in developed countries in which neither author was a resident 
in the area actually had a significantly lower proportion of findings implemented than those in 
developing countries (fig. 29). 
Fig 29. Proportion of findings implemented in developed and developing countries when there was no 
resident author. Developed countries had lower levels of implementation. 
 
To further emphasise the reason for three way interaction, fig. 30 shows that NGO or 
government affiliations are likely to result in implementation of findings in developed 
countries when co-authors are resident to the country of study, but not in developing countries. 

---

56 
Fig 30. Levels of implementation according to author affiliations when co-authors are resident to the 
country of study in (a) developing countries (b) developed countries. In developing countries, there is no 
difference in findings implemented dependant upon to author affiliations, but in developed countries 
NGO/Govt affiliations are an important predictor of implementation of findings. 
4.6.1 Multivariate analysis of factors influencing implementation in developing countries 
The significant difference between developed and developing countries, and the interactions 
described above suggests that different factors govern whether or not findings are 
implemented in developing countries to developed (The same variables were significant upon 
multivariate analysis of developed countries as for the whole data set and the model is 
therefore not shown).
For developing countries, as suggested by the interactions above, whether or not the author 
was academic made no difference to implementation of findings. Also, when funding bodies 
were regressed against the response variable, there was no significance for local government 
as there previously had been, and funding by international NGOS was retained in the 
minimum adequate model. Only co-authors affiliated with a local NGO had an impact, but this 
was not retained in the model (table 10).