ion in two forms, one
suitable for use in scoring, the other in heavy type to be read by the
subject.

PROCEDURE. Hand the selection to the subject, who should be seated
comfortably in a good light, and say: "_I want you to read this for me
as nicely as you can._" The subject must read aloud.

Pronounce all the words which the subject is unable to make out, not
allowing more than five seconds' hesitation in such a case.

Record all errors made in reading the selection, and the exact time. By
"error" is meant the omission, substitution, transposition, or
mispronunciation of one word.

The subject is not warned in advance that he will be asked to report
what he has read, but as soon as he has finished reading, put the
selection out of sight and say: "_Very well done. Now, I want you to
tell me what you read. Begin at the first and tell everything you can
remember._" After the subject has repeated everything he can recall and
has stopped, say: "_And what else? Can you remember any more of it?_"
Give no other aid of any kind. It is of course not permissible, when the
child stops, to prompt him with such questions as, "_And what next?
Where were the houses burned? What happened to the fireman?_" etc. The
report must be spontaneous.

Now and then, though not often, a subject hesitates or even refuses to
try, saying he is unable to do it. Perhaps he has misunderstood the
request and thinks he is expected to repeat the selection word for word,
as in the tests of memory for sentences. We urge a little and repeat:
"_Tell me in your own words all you can remember of it._" Others
misunderstand in a different way, and thinking they are expected to tell
merely what the story is about, they say: "It was about some houses that
burned." In such cases we repeat the instructions with special emphasis
on the words _all you can remember_.

SCORING. The test is passed _if the selection is read in thirty-five
seconds with not more than two errors, and if the report contains at
least eight "memories."_ By underscoring the memories correctly
reproduced, and by interlineations to show serious departures from the
text, the record can be made complete with a minimum of trouble.

The main difficulty in scoring is to decide whether a memory has been
reproduced correctly enough to be counted. Absolutely literal
reproduction is not expected. The rule is to count all memories whose
thought is reproduced with only minor changes in the wording. "