9 8

EMPLOYMENT  PSYCHOLOGY

letter  and  number-substitution  test  bears  to  typists  and
stenographers.  It  has  been  found  that  this  test  correlates
very  well  with  the  subsequent  ability  of  operators  who  are
not  expert  at  the  time  when  the  tests  are  given.  The  socalled
  general-intelligence  tests  which  have  been  applied
to  comptometer  operators  thus  far  have  not  shown  any
marked  correlation  with  their  ability  as  operators.  Since
the  work  on  which  these  operators  are  engaged  requires
intelligence  of  a  peculiar  kind  this  is  not  surprising.
After  tests  had  been  given  to  applicants  for  comptometry
  for  several  months,  the  value  of  the  tests  had  won  the
recognition  of  most  of  those  concerned.  A  certain  official,
however,  head  of  the  largest  group  of  operators,  remained
skeptical  and  would  not  admit  that  the  tests  had  been  successful ­
  in  selecting  operators  for  his  group.  In  order  to
convince  him  and  also  for  the  sake  of  making  further  improvements ­
  in  the  tests  themselves,  this  official  was  asked
to  allow  an  experiment  to  be  conducted  with  some  of  the
members  of  his  group.  He  consented  to  the  proposition
upon  two  conditions:  First,  he  was  to  be  permitted  to
select  the  girls  to  be  tested;  and  secondly,  he  was  to  keep
the  standing  of  these  girls  unknown  to  the  experimenters
until  the  latter  had  completed  their  ranking  of  the  girls
according  to  the  tests  and  were  prepared  to  submit  these
rankings  to  him  in  written  form.  These  conditions  were
naturally  acceptable,  since  it  is  situations  of  exactly  this
kind  that  employment  tests  are  calculated  to  meet.  This
official  then  selected  twelve  girls  for  the  experiment,  and
the  comptometry  tests  discussed  here  and  given  in  the
Appendix  were  given  to  them  group-wise.  When  the  results ­
  were  complete,  the  twelve  girls  were  arranged  in  one,
two,  three  order.  This  was  done  both  for  each  test  individually ­
  and  on  the  basis  of  the  tests  taken  collectively.