specially devised for testing with the Stanford
revision. It contains all the necessary printed material, including
digits, sentences, absurdities, fables, the vocabulary list, the reading
selection, the square and diamond for copying, etc., and in addition
gives with each test the standard for scoring. It is so arranged as to
afford ample room for a _verbatim_ record of all the child's responses,
and contains other features calculated to make testing easy and
accurate. Regarding purchasing of supplies see p. 141.

Besides saving valuable time, a little methodical foresight of this kind
adds to the success of the test. If the child is kept waiting, the test
loses its interest and attention strays. See to it, if possible, that no
lull occurs in the performance.

Inexperienced examiners sometimes waste time foolishly by stopping to
instruct the child on his failures. This is doubly bad, for besides
losing time it makes the child conscious of the imperfection of his
responses and creates embarrassment. Adhere to the purpose of the test,
which is to ascertain the child's intellectual level, not to instruct
him.

DESIRABLE RANGE OF TESTING. There are two considerations here of equal
importance. It is necessary to make the examination thorough, but in the
pursuit of thoroughness we must be careful not to produce fatigue or
ennui. Unless there is reason to suspect mental retardation, it is
usually best to begin with the group of tests just below the child's
age. However, if there is a failure in the tests of that group, it is
necessary to go back and try all the tests of the previous group. In
like manner the examination should be carried up the scale, until a test
group has been found in which all the tests are failed.

It must be admitted, however, that because of time limitations and
fatigue, it is not always practicable to adhere to this ideal of
thoroughness. In testing normal children, little error will result if we
go back no farther than the year which yielded only one failure, and if
we stop with the year in which there was only one success. _This is the
lowest permissible limit of thoroughness._ Defectives are more uneven
mentally than normal children, and therefore scatter their successes and
failures over a wider range. With such subjects it is absolutely
imperative that the test be thorough.

In the case of defectives it is sometimes necessary to begin with random
testing, until a rough idea is gained of the