STENOGRAPHERS,  TYPISTS,  AND  COMPTOMETRISTS  89
orie  hundred  and  forty  comptometrists,  and  finally,  to  more
than  one  hundred  and  twenty  candidates  for  comptometry.
^hat  is,  more  than  one  thousand  people  were  tested  and
ftote  than  five  thousand  tests  were  given.  The  tests
selected  on  the  basis  of  these  experiments  are  those  which
showed  the  highest  and  most  consistent  agreement  with
the  abilities  of  those  examined.  By  no  means  comprehensive ­
  or  final,  they  have  nevertheless  proved  them-Se
 Jves  practical  guides  in  the  selection  and  grading  of  applicants ­
  for  the  kinds  of  office-machine  work  mentioned.
tost  of  these  tests  are  given  in  the  Appendix,  with  direct'ons
  for  use,  under  the  names  of  the  work  to  which  they
apply.

typists  and  dictating-machine  operators
Obviously,  the  most  important  test  which  can  be  given
to  a  typist  is  a  test  in  the  actual  work  of  typing  a  given
copy  or  f orm<  Several  tests  were  selected  or  devised  for
.  s  purpose  and  tried  out  on  a  number  of  typists  with  a
g lew  of  discovering  which  tests  were  most  significant.
°rne  very  important  points  were  revealed  by  these  pre-"ninary
  trials.  In  the  first  place,  it  soon  became  apparetlt
  that  the  tests  must  be  given  on  a  machine  of  the  type
to  w hich  the  applicant  had  become  accustomed  or  on
hich  she  had  been  trained.  A  slight  difference  in  ma-^
  'nes  was  frequently  enough  to  throw  a  typist  out  of
er  stride  and  cause  her  to  make  an  unusual  number  of
typographical  errors,  errors  which  were  obviously  due  to
^familiarity  with  the  machine.  The  solution  of  this
Acuity  was  to  allow  each  applicant  to  take  the  typing
test  on  the  machine  of  her  choice.  However,  where  this
^ as  ^practicable,  a  standard  machine  was  chosen,  and
e  a pplicant  given  sufficient  time  to  adjust  it  to  her  use.