74 
EMPLOYMENT PSYCHOLOGY 
It will be seen that test number 51 is the only test which 
shows a significant correlation. 
In order to verify the results obtained and also because 
it was considered that not enough men had been examined 
to make the results very reliable, another set of examina 
tions was given about six months later. In this section 
of the experiment fifty-one finished gun assemblers and 
twenty-six action assemblers were given the spatial per 
ception test number 51 and in addition the Stenquist 
mechanical test already described. It had been hoped that 
by this time it would be possible to obtain piece-work 
records showing the comparative productiveness of the 
various men. However, although the men were on piece 
work, each man was limited to a certain number of pieces 
a day; therefore, it was again impossible to obtain pro 
duction records showing the comparative ability of the 
men. Again it was necessary to resort to the opinion of 
the foreman. On this occasion both the immediate over 
seer of the men and the foreman were asked to rank the 
men independently in five groups according to their opin 
ion of the productive ability of their workers. The rank 
ing of the men in the tests was then mathematically cor 
related with the ranking of the overseer and the foreman 
taken separately and with the combined ranking of both 
men. The results are given in the table on page 75. 
It will be seen at once that the correlations for test 51 are 
not as high as was the correlation found in the first part 
of the experiment, though high enough to encourage 
further experiments in the same direction. The Stenquist 
test shows a slightly higher correlation, although it, too, 
is not high enough to justify its use as a criterion in the 
selection of new assemblers. The correlation between the 
opinions of the foreman and the overseer is very high,