Increasing the Generalizability of Qualitative Research, J W Schofield, 1989, Vol 1 Reader, p91

What is, may be and could be - solving the riddle of generalisability.

At last a paper that says what it says it will say, says it succinctly and imparts a genuine new and useful understanding of educational qualitative research. The summary says it all really: the paper itself is full of reasoned argument and examples drawn from the author's own experience.

Janet Ward Schofield argues that there are two questions one should ask if one wants ones research to be transferable to others:

1. What do we want to generalise to?
2. How do we design to maximise generalisability?

If Schofield is right, generalisability "is best thought of as as a matter of the 'fit' between the situations studied and others ..." To be able to judge how fit a study is, one needs information about the studies' concepts and conclusion - a thick description.

There are 3 useful targets for generalisation:

1. what is (e.g. accepted best practice)
2. what may be (e.g. unusual, but possibly good practice)
3. what could be (e.g. practice that could take off, but might not)

I would re-write these as more the more mnemonic present, possible and potential practices.

An enjoyable paper with sage advice.

NB p107 has an error, penultimate paragraph where "may be" should be "could be".