The records at the Institute for Juvenile Research seem to be
admirably suited for the present investigation. There are on file
more than ten thousand case records which have been accumu-
lated during the last ten years on children varying in age from
one year to twenty-one years. One of the routine requirements
for practically every case study is a Stanford-Binet examination,
which is only occasionally varied by introducing some other re-
vision of the Binet test. All of the mental ages and intelligence
quotients that have been used in this study are based on indi-
vidual psychological examinations. No group tests have been
used for this study.

From the case history it is relatively easy to ascertain the
number of siblings for each patient and their ages. By a relatively
simple process of comparing cards in a cross index of family
names, it was also possible to pick out all families which have sent
more than one child to the Institute for examination and so to
make comparisons between all possible pairs of siblings in the
Institute records. In assigning ordinal number of birth to each
child, the question naturally arose whether miscarriages should be
counted. In many cases the records are rather complete in this
regard but in many others the information is of doubtful ac-
curacy. It was finally decided that the miscarriages would not be
taken into consideration in assigning birth number to each patient
because of the fact that all of the case records are not equally
trustworthy on this point. The ordinal number assigned to each
child takes into consideration all of the siblings who are listed in
the case history including deaths but not miscarriages.

-1-