§ 10. CO-OPERATION OF THE SUBJECT.
In our work with insane subjects we encountered many cases in which we were unable to obtain satisfactory test records owing to lack of proper co-operation. Some subjects seemed to be either too confused or too demented to be capable of understanding and following the instructions given them. Others were for one reason or another unwilling to co-operate. It is important to distinguish inability from unwillingness to co-operate, since the former indicates in itself an abnormal state of the mind, while the latter is quite often shown by normal persons.
A subject may co-operate to the extent of giving a single word in response to each stimulus word, and yet fail to co-operate in some other particulars.
He may, instead of giving the first word suggested to him by the stimulus, suppress the first word more or less systematically, and give some other word which may seem to him more appropriate. This probably occurs very often, but does not seem to render the results less serviceable for our purpose.
Further, a subject may react by words related not to the stimulus words, but to each other, thus simulating perseveration; or he may react by naming objects within reach of the senses, thus appearing to be distracted; or he may give only sound reactions.
There is, in fact, no type of pathological reactions which a normal person may not be able to produce more or less readily at will, though in the case of incoherent reactions considerable mental effort may be required, and the end may be attained only by regularly rejecting the first and some subsequent words which are suggested by the stimulus.
In view of these considerations we are led to conclude that the association test, as applied by our method, could not be relied upon as a means of detecting simulation of insanity in malingerers, criminals, and the like.
§ 11. SUMMARY.
The normal range of reaction in response to any of our stimulus words is largely confined within narrow limits.
The frequency tables compiled from test records given by one thousand normal subjects comprise over ninety per cent of the normal range in the average case.