Designing Effective Farm Practice Survey Questions.

Farm practice data is one of several information pieces required by agricultural researchers, extension specialists, and program/policy analysts to conduct effective work.  Knowing what farmers do is often the information piece most lacking.  Individual farms are complex entities, meaning that specific practices vary spatially and temporally.  This is due to variable soil and weather, different crop types, changing economic and social conditions, and other factors. 

Many sources of information about farm practices are subjective, meaning they lack the precision required for research and modelling purposes.  Some examples include farm journals and opinions of agriculture specialists who work with farmers.   Sometimes researchers use sources that provide recommended farm practice, but one cannot assume farmers are following recommendations.   A more objective source of farm practice data is a survey or questionnaire.   However, even a survey presents numerous challenges.  It is impossible to fully document what happens on complex farm operations, especially when farmers are reluctant to commit significant time to complete a survey.  For a survey to be effective it must capture the most and best information in the shortest time possible. 

This presentation will provide insights on designing effective survey questions related to soil and crop management, based on over 10 years experience with the Farm Environmental Management Survey (FEMS) in Canada.  The FEMS is a collaborative initiative of Agriculture and Agri-Food Canada and Statistics Canada.   Recently, FEMS has initiated a sector approach whereby separate questionnaires are being developed for annual field crops, perennial forages, horticulture crops, dairy, beef, pig, and poultry,.  This approach enables better focus on practices related to specific production systems, and facilitates greater opportunity to measure agronomic and environmental performance of specific sectors.  Various methods will be presented for designing questions and answer options to improve efficiencies while maintaining accuracy.  A few examples will illustrate how the wording of a question is critical for making it easy for a farmer to respond and at the same time providing data that is useful to the researcher or modeller.  Finally, the impacts of various question/answer options on required data manipulation and analysis to enable proper interpretation will also be discussed.