In 1910, Mercer and Hall conducted a wheat yield experiment at Rothamsted
Experimental Station in Great Britain. The experiment, a uniformity trial, consisted of giving a 20 by 25
lattice of plots the same treatment (presumably fertilizer, water, etc.). Visually, we identify the histogram's
shape as the familiar bell-shaped curve, indicating the nearly normal distribution of the 500 wheat yield measurements.
The mean (3.95), median (3.94) and mode (3.97) are nearly equal, and the the skewness (.036) and kurtosis (-.254)
are close to zero, all providing evidence that the distribution is close to normal. The exact size of the
plots from the original data set seems to be unknown, although some researchers have used 3.30 meters east to west,
and 2.51 meters north to south. To view the spatially arranged data, download the Rich Text Formatted document
above. Methods to analyze spatial statistics of these data are not discussed here; an excellent source for
this topic is the text by Young and Young listed below.
The Rothamsted Experimental Station, in Great
Britain, is part of the the Institute of Arable Crops Research (IACR).
Rothamsted, founded in 1843, claims to be "the oldest agricultural
research station in the world". In 1843 they started a
series of long-term field experiments whose interest was to measure
the effect on crop yields of inorganic and organic fertilizers.
For more information, visit the website of IACR.
Unfortunately, no data from the long-term experiments are posted on
the website.
Data cited in: Young, L.J. and J.H. Young, 1998, Statistical Ecology, A Population Perspective,
Kluwer Academic Publishers, p. 232.
Original Data from: Mercer, W.B., and A.D. Hall. 1911.
The experimental error of field trials. Journal of Agricultural
Science (Cambridge) 4: 107-132