Martyn Hammersley is my favourite methodologist in the social sciences. His views are pragmatic without being superficial, simplistic or favouring certain tendencies or approaches.
Recently, he published the paper What is ethnography? Can it survive? Should it?. By the end of the paper, the author proposes a table showing the options for the design of research in social sciences and this table is complemented by a diagram attached in the end of the paper. Because the diagram was poorly designed, I contacted the author asking for clarifications. Martyn kindly answered and even provided some complementary notes on the table and diagram.
As a result, I compiled all the information in the diagram below.
The svg file can be downloaded here.
I find it extremely useful to clarify ideas related to research design and choices that a researcher has to make during this process.
Some choices in one stage may limit, or even determine, the options available in the next stages. So, it is not possible (or it just does not make sense) to combine all the possibilities freely.
The table is not a suggestion of order in terms of when the choices should be made.
The "Choice of cases" aspect is discussed in the chapter So, what are case studies? in the book What's wrong with ethnography?.
Hammersley recognizes that the item Data Analysis deserves more details, but he had " not even attempted to map the dimensions involved [t]here".
The content of the post was totally based on papers by Martyn Hammersley and should by treated/referred as usual in academic papers. When I contacted him, he was clear that I could use the extra material for whatever end I wanted. However, the diagram above is my reading of his work and was not extracted directly from his papers.