39 6 
APPENDIX 
Teachers’ College, are a great convenience for the beginner. 
A comprehensive treatment may be found in E. L. Thorndike’s 
“Mental and Social Measurements” and a very concise pres 
entation in Guy M. Whipple’s “Manual of Mental and Physi 
cal Tests ”, Vol. I. The practical psychologist will find that 
the slightly greater accuracy of the Pearson formula is more 
than compensated for by the greater flexibility and simplicity 
of the Spearman formula. It takes about six times longer to 
work out the Pearson formula than the Spearman formula 
(differences squared). In the treatment of employment tests 
it is much more important to obtain a large number of corre 
lations than to spend time trying to increase the refinement of 
a few. Two indices in which there is an inaccuracy of even 5 
points have a greater practical significance than one correla 
tion which is absolutely accurate. 
Where groups of more than 20 rankings are to be compared, 
the Spearman “footrule” is sufficiently accurate, and extremely 
convenient. Extensive use of this formula was made in the 
course of the experiments described here. Where “ n” was less 
than 20, the longer Spearman formula was used. 
Probable errors are not given in connection with the indices 
of correlation in the body of the book for two reasons: first, 
because they would confuse the ordinary reader; secondly, the 
practical psychologist can tell almost at a glance what the 
probable error is. Knowing the number of individuals and the 
size of the correlation, the P. E. can easily be deduced. 
B. GENERAL PROCEDURE 
When an individual applies for work, his application is taken 
in the employment office and the information placed on a per 
manent employment record card. This card is then taken into 
the psychological examination room where it serves as an an 
nouncement of the applicant’s readiness to be examined. It 
also tells the examiner the kind of work for which the applicant 
is asking and gives him certain other essential information.