Information for G01697
PLA Notes CD-ROM 1988-2001 4 How big should on-farm trials be and how many plots should be measured?
Document begins: PLA Notes CD-ROM 19882001 4 How big should on-farm trials be and how many plots should be measured? William J. Fielding and Janet Riley Introduction On-farm trials are often conducted where only limited areas of land are available and where Traditional agricultural research has been the there is the likelihood of large differences in preserve of scientists who have acted on behalf production from different sections of the field. of farmers. This research resulted in many The usual approach to research comparing measurements of great precision which can be different technologies would require many analysed according to standard procedures. harvest areas to be planted to get precise (if inaccurate) estimates of yield. Large harvest However, no matter how good the science might areas would be required to get accurate yields be or how persistent the extension efforts, it has for economic analyses and assessments of inputs been found repeatedly that technological advances will not be adopted unless they are (particularly labour). acceptable to farmers. For example, in Jamaica Yield measurement the recommended practices for the control of sweet potato weevil have been known, and extended, for 100 years but never adopted by Current literature indicates that to assess ...