2  6

EMPLOYMENT  PSYCHOLOGY

eyes  or  letting  up  of  the  attention  is  likely  to  allow  some
bad  shells  to  slip  by  or  to  lengthen  the  operation.
This  analysis  having  been  made,  the  next  step  was
to  find  tests  which  would  be  likely  to  detect  the  presence
of  these  qualities.  Sixteen  different  tests  were  chosen
and  prepared  for  preliminary  trials.  It  is  not  possible
or  necessary  to  go  into  detail  on  all  these  tests  at  this
point.  However,  each  test  was  tried  out  on  a  large  enough
number  of  girls  to  give  a  good  indication  as  to  whether
it  was  likely  to  prove  significant.  A  good  deal  of  time
was  required  for  this  part  of  the  experiment,  but  eventually ­
  a  set  of  eight  tests  was  selected  for  the  body  of  the
experiment.  These  eight  tests  will  be  described  later;  but
before  describing  them,  one  or  two  preliminary  problems ­
  will  be  of  interest.
An  important  question  to  be  settled  was  the  question
as  to  just  how  these  tests  should  be  given.  The  rooms  in
which  the  girls  were  at  work  were  very  noisy,  due  to  constantly ­
  running  machines  and  the  handling  of  thousands
of  brass  shells.  Should  the  girls  be  tested  in  this  noisy
atmosphere  or  should  they  be  taken  off  to  some  quiet
place,  free  from  any  possible  disturbance?  In  the  end  it
was  decided  to  give  the  tests  in  the  workroom,  on  the
supposition  that  if  the  subjects  were  left  in  their  regular
environment,  they  would  be  more  likely  to  show  characteristic ­
  results  in  their  performance  in  the  tests.  A  girl
suddenly  taken  out  of  a  noisy  shop  to  which  she  had  become ­
  accustomed  and  into  a  soundless  room  might  feel
just  as  strange  as  a  girl  taken  from  a  quiet  place  into  a
very  noisy  one.  A  small  room  was  screened  off  on  one
side,  and  this  served  as  a  place  in  which  the  tests  could  be
given  with  comparative  freedom  from  intrusion.
Some  difficulty  had  been  anticipated  in  putting  the