28 
BULLETIN 1480, U. S. DEPARTMENT OF AGPJCULirBE 
two dimensions — quality and location — whereas prices for a season 
would be subject to variations in a third dimension — time. The rela- 
tive probable errors of prices for crop-reporting districts average 
about 1 per cent. The relative probable error of the price for the 
State as a whole is very small — 0.312 per cent — because of the large 
number of reports, namely 743. 
A highly interesting comparison was made of the relative probable 
errors of the straight (unweighted) average and the weighted 
average of Kansas wheat prices during a crop year. 1921^25. As 
stated previously (p. 26). the probable error of the weighted average 
(3. p. 317) is greater than that of the unweighted average. This ad- 
vantage of the unweighted average is at least partly offset by the 
statistical device of stratification of the sample. The relative prob- 
able error of the straight average was 0.312 per cent and the relative 
probable error of the weighted average, weighted by the wheat acre- 
ages in the crop-reporting districts, was 0.330 per cent. 
CORX 
Table 10 shows a comparison of the corn-price samples from sev- 
eral States for 
(as reported b\ 
Dakota. 
the 15th of the month, and December 
prices 
crop reporters) for Maryland, Iowa, and Xorth 
'able 10. — Farm prices of corn: Selected illustrations of size of sample, meas- 
ures of dispersion, end probable error 
[Per bushel] 
State and date 
' i - 
Number £*»_ 
of reports, <£}{* 
1 mean) 
Standard 
Coeffi- 
devia- 
cient of 
tion of 
varia- 
bility 
• 
Per cent 
9.3 
9.4 
5.8 
5.6 
5.1 
9.1 
18.7 
18.8 
14.5 
18.0 
7.9 
11.0 
16.4 
22.0 
5.9 
10.4 
13.2 
23.3 
Probable 
error 
of the 
average 
price or 
mean 
; Pour 
Relative ; times 
probable relative 
error ; probable 
I error.* 
Iowa: 
May, 1925 
June, 1925 
May. 1926 
Georgia, [November, 1920 
Pennsylvania, November, 1926 . 
Maryland: 
Mar. 15, 1926 
Dec. 1,1925 - 1 
Iowa. Dec. 1. 1925 * 
North Dakota. Dec. 1. 1925 -> 
94 
S3 
1S4 
55 
43 
3S 
115 
341 | 
75 I 
Cents 
98. ■ 
103.3 
56.4 
99.3 
80.7 
70.9 
74.4 
56.9 
56.6 
0.63 , 
.2o 
1.70 
1.49 
.87 
1.03 
.22 
Lll 
Per cent 
0.6 
.4 
.4 
1.7 
1.9 
1.2 
1.3 
.4 
2.0 
Per cent 
2.4 
1.6 
1.6 
6.8 
4.S 
5.2 
1.6 
8.0 
1 The probabilities are ninety-nine out nf one hundred that the average of a much larger sample collected 
in the same way an i . rime would not vary from this average by more than four times the probable 
error. 
he Dec. 1 prices were reported by crop reporters and not by the regular price reporters, who report 
on the 15th of each month. 
Prices of corn in p. surplus-producing State like Iowa, have only 
about half the variability of prices of corn in a deficit corn-produc- 
ing State like Georgia or Pennsylvania. The Iowa farm price of 
corn for May. 1926. was 56.4 cents as compared with 103.3 cents 
in June. 1925. and the standard deviations of the prices received 
were 5.1 and 5.S cents. With the much larger sample in May. 192(5. 
184 reports as compared with S3 in June, 1925, the relative probable 
errors of the averages were 0.26 cent and 0.43 cent. But with the 
average price in May, 1926. much smaller than in June. 1925. the 
relative probable error was the same for both averages (0.4 per cent). 
