In quantitative empirical research, what makes a survey good? Well, not perfect, but optimally good – that is, its questions ask exactly what we want them to ask, and the response rate is as high as possible? There are no recipes, but sometimes the survey method, with all its epistemological and methodological flaws*, is still the best method to use for our particular study.
So, if you are using surveys, here is a neat table which summarises some good tips for writing a survey. It lists several criteria for the questions: they have to be simple, specific, individual, exhaustive, neutral and balanced; and sometimes it is a good idea to make questions optional:
* Here is a random and by no means exhaustive selection of sources for a first look into these issues (a lot has been written).
For a general discussion see e.g. Marsh 1979 and Bryman 1984 (both are old but good); or any of the many editions of Bryman’s book ‘Research Methods’ (here is a scanned version of the chapter on quantitative research from the 2004 edition).
For a feminist discussion of the merits of qualitative/quantitative research see e.g. Westmarland 2001.
For problems of response, see e.g. this summary.