EMPLOYMENT  PSYCHOLOGY

5 2

service  could  therefore  be  estimated,  the  results  were  as
follows:
Length  of  Service

Inspectors  recommended  by  the  tests  9.56  weeks
“  not  recommended  but  hired  1.05  weeks

In  other  words,  the  average  term  of  those  who  were  recommended ­
  and  hired  on  the  basis  of  the  tests  was  almost  ten
times  as  long  as  the  average  term  of  those  who  were  hired
contrary  to  the  evidence  of  the  tests.  This  is  really  a
striking  commentary  on  the  results  of  these  tests  in  so  far
as  they  affected  the  turnover.
To  summarize,  individual  cases  showed  that  the  personal ­
  judgments  of  the  experimenter,  foreman,  and  instructors ­
  were  not  nearly  so  reliable  as  the  evidence  of
the  tests.  A  large  number  of  cases,  represented  by  curves,
showed  that  the  tests  made  possible  at  once  a  separation
between  piece-workers  and  day-workers  which  was  quite
beyond  the  scope  of  the  ordinary  employment  methods.
This  graphic  representation  showed  also  how  the  standards
which  made  possible  the  desired  distinction  between  applicants ­
  was  found.  The  careful  steps  by  which  these  tests
and  standards  were  introduced  into  the  employment
routine  were  traced.  Finally  the  concrete  results  of  applying ­
  these  tests  in  the  employment  office  were  presented.
It  was  found  (1)  that  the  verdict  of  the  foreman  tended
to  bear  out  that  of  the  tests;  (2)  that  of  a  certain  group
studied  intensively,  ninety-four  per  cent  of  the  successful ­
  workers  were  above  the  standard  in  the  tests,  and  all
failures  were  failures  also  in  the  tests;  (3)  that  those  who
passed  the  tests  worked  almost  ten  times  as  long  as  those
who  did  not.