I am going to prepare a survey for farmers in East Africa. The data aquired I want to upscale to village scale and from there to regional scale. I am searching established methods for this upscaling process and the criteria I have to look for.
I am thankful for your answers.
This is probably a frustratingly vague answer, but it depends on the type of survey you are doing and the data you're collecting. The key to any type of scaling is to compare relevant characteristics of your sample to those of the target populations. The more similar they are, the more confidence you can have in your extrapolation. The two broadest categorizations you should think about are:
1) quantitative vs. qualitative data
2) something I call "hard vs. soft" information. Meaning that if you're looking at more or less empirical data--for example, crop yields, number or type of crop pests observed in a season, whether crop yields were lower than in past years, etc. (these are pure speculation)--you will handle extrapolation differently than if you are looking at more subjective data--economic, sociological, respondents' attitudes or perceptions, etc.
If you're looking at empirical, quantitative data, it's relatively straightforward to find statistical methods that will mitigate the effects of biases, disparate sample sizes, and the like. On the other end of the scale, if you've got qualitative, subjective data, you may not be able to put numbers on it easily. In that case you would have to research demographic and cultural characteristics of the region to make the call as to whether your information from farmers can be applied more broadly. This is something that comes up in epidemiology all the time.
So the short version of my response is, it's hard to tell without knowing more about your study.
you can follow cross sectional survey by purposive sampling method !
For any kind of survey better to say as bench mark survey it is necessary to know the criteria to be selected from farmers point of view. After that you have to prepare a structured interview schedule by which you can directly communicate individual farmers for getting first hand information. Another method to extract actual information is to do PRA (Participatory Rural Appraisal). There are different methods of PRA which you will get in available literature like van diagram, transect walk, livestock analysis and so on. I think first you will design your work and accordingly you can implement it in village level and then afterwards you can implement in regional level through some Farmers club or such type of small innovative groups created in each of selected villages.
- Badges
- Science topic
- Similar topics
- Agricultural Science