The term 'failure' is employed in this study to signify the non-passing
of a pupil in any semester-subject of his school work. The school
decision is not questioned in the matter of a recorded failure. And
although it is usually understood to negate "ability plus
accomplishment," it may, and undoubtedly does, at times imply other
meanings, such as a punitive mark, a teacher's prejudice, or a deferred
judgment. The mark may at times tell more about the teacher who gave it
than about the pupil who received it. These peculiarities of the
individual teacher or pupil are pretty well compensated for by the
large number of teachers and of pupils involved. The decisive factor in
this matter is that the school refuses to grant credit for the work
pursued. The failure for a semester seems to be a more adaptable unit
in this connection than the subject-failure for a year. However, it
necessitates the treatment of the subject-failure for a year as
equivalent to a failure for each of the two semesters. Two of the
schools involved in this study (comprising about 11 per cent of the
pupils) recorded grades only at the end of the year. It is quite
probable that the marking by semesters would actually have increased
the number of failures in these schools, as there are many teachers who
confess that they are less willing to make a pupil repeat a year than a
semester.

By employing this unit of failure, the failures in the different
subjects are regarded as comparable. Since only the academic and
commercial subjects are considered, and since they are almost uniformly
scheduled for four or five hours a week, the failures will seem to be
of something near equal gravity and to represent a similar amount of
non-performance or of unsatisfactory results. There were also a few
failures included here for those subjects which had only three hours a