The Movements of Countrymen 

 Table 17— Continued 



in 



(a) Products and squares of difference are used to correct the product and 

 square columns after dividing the totals by the number of cases (100.) Inas- 

 much as 000 is omitted from the product and square columns, after dividing by 

 100, only is omitted. The proper correction is indicated in the last line by 

 dividing the sums of the products and squares by 100, i. e. by adding to the 

 totals which omit 000, and subtracting the products and squares of the dif- 

 ferences between guessed and true averages. 



onrl pinning tVmm i-mrfcf fhr fignrfl t n fe r mnrrtH' 



For fuller explanation of Guessed Average method, see appendix. 



Figures computed from U. S. Census, Negro Population in the U. S. Table 

 II and 73, Census of 1900, Vol. V Agriculture, Table 10. 



The counties grouped at the end of the table make up areas in which new 

 counties were created between 1900 and 1910, see foot note page 112. The 

 number of counties indicated in each group is the number which composed the 

 area in 1900. Their increases have been added to the population of the new 

 counties created in the area. In the mathematical operations this group increase 

 was divided by the number of counties in order that each county might be treated 

 as a single case on the same basis with other counties in the table. For instance, 

 in Group I, the group population increase was 911, of which 455.5 was assigned 

 arbitrarily to each of the two counties which were in this area in 1900. Subtract 

 the guessed average from 455.5, i. e. 455.5 — 700 and the result is — 244.5, which 

 is the deviation of the population increase of each of the two counties. Multiply 

 this by two and the result, — 489, is the deviation of the group from the popula- 

 tion average. Multiply the square of — 244.5 by two and the result, 120,254 is 

 the Y square for the group. Multiply — 244.5 by — 192.5, which is the deviation 

 of each county from the average increase in farms, and the result, 47,066, when 

 multiplied by two gives 94,132, which is the product of the two deviations for the 

 group. This process has been followed with the five counties in Group II, the 

 six in Group III, the two in Group IV and the two in Group V, thus giving to 

 each group a weight corresponding to the number of counties which composed 

 it in 1900. 



