Abstract / Description of output
In this paper the application of nonlinear and expert modelling approaches to social science data are discussed and illustrated with examples from the Edinburgh Survey of Decision-Making on Farms. In general it is found that nonlinear models do not give appreciably enhanced goodness-of-Jit compared to linear models. The reasons for this, and the relatively poor fit found for an expert model are discussed; it is pointed out that the strength of nonlinear and expert models is more likely to be associated with the understanding derived from them than from goodness-of-fit. Some general conclusions about the success of the survey and the models obtained from the data are discussed and proposals for possible new approaches to surveys of farmers put forward.
Original language | English |
---|---|
Pages (from-to) | 225-241 |
Number of pages | 17 |
Journal | Agricultural systems |
Volume | 58 |
Issue number | 2 |
DOIs | |
Publication status | Published - Oct 1998 |
Keywords / Materials (for Non-textual outputs)
- DECISION-MAKING
- LINEAR-MODELS