RELIABILITY AND ADEQUACY OF FARM-PRICE DATA 43 
SUMMARY 
Farm prices of cotton, wheat, and flax have small coefficients of 
variability in most States. Prices of corn, oats, hogs, veal calves, 
lambs, butter, butterfat, and eggs also show small variability in 
surplus-producing States. In most States the number of reports 
received as to the prices of these products is sufficient to render the 
monthly State price averages reasonably stable and reliable. 8 
Speaking in terms of probability, the chances are ninety-nine out 
one hundred or there is practical certainty that the average of a 
much larger sample taken at the same time and in the same way 
would fall within a range of from 1 to 5 per cent of the average 
obtained by the present sample. 
With the prices of the remaining farm products it is only in those 
States where an unusually large number of reports are received that 
the size of the sample is sufficient to offset the higher variability and 
reduce the probable error to a point where four times the relative 
probable error is much below 10 per cent of the average price. 
Apples are about the only farm product where the variability in 
the prices of which is so high that four times the relative probable 
error is likely to exceed 20 per cent. 
Generally speaking, the December 1 prices of crops and the Jan- 
uary 1 values of livestock are based on so many more reports than 
the State monthly average prices, especially those of minor farm 
products, that they are much more reliable than are the monthly 
prices. It is not unusual, however, for a December 1 price or January 
1 value sample to show a greater degree of variability than a cor- 
responding monthly price. An additional safe guard at the present 
time is the fact that the December 1 price and the January 1 values 
are also obtained from another list of crop reporters who report 
to the State statistician. The results from both samples are combined 
to obtain the final figure. 
' A COMPARISON OF STATE FARM PRICES 
The preceding analysis has been based entirely on the sample of 
price reports received for a given month by States. If the prices 
as reported each month for the major farm products in the larger 
States are reasonably stable and dependable from month to month, 
how do the price series in two different States compare over a period 
of several months or years? Do the prices in both States tend to 
move together each month or do they move in opposite directions at 
times? If they tend to move together, confidence in the reliability 
of the price series is increased. 
8 These monthly prices are subject to defects in representativeness and to errors in 
computation, especially when the work on monthly prices is pushed rapidly each month 
in order to make the 15th-of-the-month prices available at the earliest possible moment. 
In months when the crop reports do not interfere the price report has been completed by 
the 25th or 27th of the month. As the price series for individual States are used in re- 
search problems, prices for a given month may appear to be inconsistent with the series. 
The Bureau of Agricultural Economics will appreciate being .told of these inaccuracies, 
that eventually a revised edition of these prices may be published containing all necessary 
corrections. Farm prices should not be used as a basis for measuring the spread between 
farm prices and market prices, except where the State represents definitely a surplus- 
producing, commercialized area of production, as the prices from both deficient and 
surplus-producing areas are contained in an average farm price for a particular State. 
