THE TECHNIQUE OF GIVING TESTS 167 
■will become increasingly necessary to prevent duplication. 
A person, making application for a job a few months after 
having been examined and rejected for the same position, 
is likely to do much better work in the tests. If a second 
and even a third trial is allowed, the essential nature and 
meaning of the tests will be radically changed. There 
fore, some definite decision must be made on the course 
to be pursued in such a contingency. Should a second or 
third trial be allowed, and should an individual be graded 
on the basis of his improvement during the interval? The 
answer to this question must be Yes and No. As will be 
pointed out in greater detail further on, tests are of two 
general kinds: first, tests designed to bring out an in 
dividual’s innate ability; and secondly, tests designed to 
make clear his degree of training, education, and expe 
rience. As instances of the first type, we may mention 
the cancellation test, the form-board test 51, and the 
tachistoscope tests. Although training does play a part 
m the performance of these tests, nevertheless native 
quickness and alertness are the most important factors. 
As examples of the second type we may name the multi 
plication and division tests, the spelling test, and the 
dictation tests. These tests are intended to measure the 
degree of education which has been secured by the appli 
cant. Where tests to determine native ability have been 
given, no second trial should be allowed, except in so far as 
the second result is used to show the effects of practice. 
Where the object is to discover the degree of education, 
a uy number of trials should be allowed. A stenographer, 
for instance, ought to be allowed a second trial after an 
mterval of two weeks or a month, because, during that 
time, she may have learned much and acquired a decided 
mcrease in speed. However, in order to avoid deception