The Relationship Between Yields on Farms and in Experiments

Abstract

A definite relationship appears to exist between the yields obtained on farms and in experiments. For crops the relationship is curvilinear and for animal products it is linear. In both cases the ratio of average farm yields to experimental yields decreases as experimental yields increase. The ratio between average farm and average experimental yields over a period of years decreases as the area of crops increases and varies with the type of animal product.

The Application of Experimental Research to Farm Production

“…It is the purpose of this article to demonstrate differences between experimental and commercial yields and to offer some explanations within the terms of classical production theory. We have previously made comparisons between farm and experimental yields for a number of crop and animal products, using a combination of cross-sectional and time-series data [1,2]. Only comparisons of crop yields based on time-series data will be referred to here, though the general conclusions can be applied also to animal products.…”

Economic effects of alternate growth path, time of calving and breed type combinations across southern Australian beef cattle environments: feedlot finishing at the New South Wales experimental site

“…The limitations of this methodological approach to extrapolation of trial data to farm-level analyses can be addressed to some extent through the appropriate validation of the model used and the use of sensitivity analyses of key assumptions (Dillon and Anderson 1990). See also the discussion in Davidson and Martin (1965) on this topic.…”

Untitled

“…Same ResuZts from a Study Zone A casual analysis of the data from the study area' led us to suspect that the farmers did not in fact take the restriction on fertilizer into 6 We have used a uniform prior distribution (the usual model in a situation of ignorance [ 10, p. 461 ) for expository convenience. The quaIitative essence of this argument holds for any diffuse prior probability distribution where the random variable e is bounded by €2.…”

“…The traditional wisdom is that such rates will exceed the optma1 levels for risk-averse farmers; 'but even if they were risk-neutral, the more timely and controlled nature of experimental operations would lead to recommendations which overstate the optimal levels for farm (and farmer) conditions [6]. Thus, inherent biases exist toward requiring a fertilizer rate greater than the unconstrained optimal rate.…”