270 Folkmar—The Duration of School Attendance. 
Corrections for Death .—Another factor should be given con¬ 
sideration in this inductive method, namely, the per cents that 
drop out through death. What correction, if any, should be 
made in the table for this cause? A table of the death rate 
per thousand will make it evident that the correction would not 
be an appreciable quantity. We may take, for instance, the 
English Life Table No. Ill, which is a standard authority in 
life insurance 1 and we shall find that during the school age, 7 
to 18 years, less than one per cent, drop out in any year on ac¬ 
count of death, (more exactly, from .9 per cent, of the first grade 
enrollment in the first year to .006 per cent, of it in the twelfth 
grade) and putting that small correction beside the large per 
cents, 43, 54, 66, and the rest in Table XVI, it is evident that 
the correction is of no practical importance. 
Inductive Method in Milwaukee .—The Milwaukee statistics can¬ 
not be used directly and with entire satisfaction in this in¬ 
ductive method, for the actual enrollment of the grades has never 
been printed. The “average enrollment ” which we have in the 
reports is a different matter. Still, if it is necessary, the act¬ 
ual enrollment might be estimated on the basis of the average 
enrollment. This will hardly be necessary after having the 
Chicago results, for we may see at once from the tables of 
average enrollments that the enrollments stand in almost ex¬ 
actly the same ratio throughout the tables as do the Chicago en¬ 
rollments; and we might get the per cents that dropped out in 
each grade directly from them — but with the understanding 
that the per cents are much too large. We are unable to reduce 
the higher grades for increase of population. The average en¬ 
rollment, therefore, in the higher grades is unduly large as 
compared with that in the primary grades; for pupils attend 
more regularly in the higher than in the lower grades. Com- 
1 Table XVIII. Influence of Death upon Duration of School At¬ 
tendance, p. 299. From the Insurance Cyclopedia , by C. Walford, 
Vol. II, pp. 528-32. Since the per cents derived from the English tables 
are computed upon the population of each age taken as a base, they must 
be recomputed (see column VI) upon the first grade as a base, to be usable 
in my tables. Column IV is taken from Table XVI. Column VI is de¬ 
rived from columns III and V. 
