264 Folkmar—The Duration of School Attendance. 
Chicago tables. 1 Comparing the second of these tables with the 
corresponding Chicago table (IY, p. 285), it will be seen that the 
averages do not vary one and one-half per cent, in the two tables 
except in the first grade, which is three per cent, larger in the 
Milwaukee table. This slight excess is clearly due to the fact 
that we are dealing with true enrollment figures in the Chi¬ 
cago table but with “average” enrollments in the Milwaukee 
table. 
Errors in Deductive Method .— But, however accurate and sim¬ 
ple this method may be in reaching the per cent, that normally 
drops out at each grade, it must be justified by comparison 
with other methods. There are errors in it not easily elimin¬ 
ated; some of which would make the per cents too large, while 
other errors would make the per cents too small. Now, if we 
apply our corrections, for example, to the second grade (Table 
III or VJ, Per Cent, in each Grade), it is evident that any error 
which makes it too small will make the per cents in the gram¬ 
mar grades too large, and vice versa; also that if the second 
grade per cent, is too small in this table, it is too small in the 
table derived from it, the first grade remaining unchanged. 
(Table IV or VII, Per Cents that Drop Out.) 
One of the two largest errors makes the final second grade 
per cent, (the 50 percent, that dropped out) too small. It arises 
from the fact that the base is too large. If the base were smaller 
the second grade per cents in both tables, as is shown above, 
would be larger. But the base should be smaller; for the grand 
total of the fifteen years’ enrollment, one and one-half millions 
(compare Tables I and IX), is evidently not pupils, but pupil- 
grades, or enrollments; that is, the number of enrollments that 
would appear on the books in all the grades. Since most pupils 
have been enrolled in more than one grade, this number is three 
or four times the number of pupils. 2 The number of pupils 
1 Table VI. Same as Table V, With Corrections for Kindergarten 
and High School, p. 287. Table VII. Per Cent, of Pupils that do 
not go beyond the Grades Named (Milwaukee), p. 288. See, also, 
“Diagram A,” facing p. 262. 
2 See discussion of the average number of grades each pupil has at¬ 
tended, p. 272. 
