400

APPENDIX

confusion,  a  method  of  rating  which  applies  equally  well  to  all
tests  was  devised.
This  method  starts  by  taking  the  perfect  completion  of  a
certain  number  of  units  in  a  certain  time  as  a  point  of  reference.
For  instance,  in  test  number  2,  sorting  the  50  cards  perfectly
in  35  seconds  is  taken  as  the  point  of  reference.  Then,  35  seconds ­
  divided  by  whatever  time  the  individual  takes  will  be
the  rating  for  the  individual.  For  instance,  an  individual  who
does  the  test  in  35  seconds  will  be  rated  35  -5-  35  or  1.00.  One
who  does  it  in  40  seconds  will  have  35  -f-  40  or  .87.  If  anybody
does  it  in  30  seconds  the  rating  will  be  35  -5-  30  or  1.20.  This
is  higher  than  1.00.  However,  by  selecting  as  our  reference
time  a  record  in  the  test  which  only  the  very  best  individuals
can  reach,  the  subsequent  ratings  will  approach,  but  seldom
pass  1.00.
In  many  tests,  however,  every  subject  is  allowed  only  a  certain ­
  length  of  time.  Therefore,  since  the  time  is  always  the
same,  it  becomes  necessary  to  divide  the  number  of  units  completed ­
  by  the  number  of  units  called  for  by  the  point  of  reference
(RU).  E.  g.,  test  8,  120  seconds  is  the  time.  The  point  of
reference  is  the  completion  of  70  numbers  in  that  time.  If
only  60  are  completed,  the  rating  in  the  test  is  60  -5-  70  or  .86.

If  the  70  are  completed,  the  rating  is  —  or  1.00.
70
The  next  step  is  to  compensate  for  mistakes.  In  tests  like
8,  where  the  time  is  always  the  same,  it  is  only  necessary  to
subtract  the  number  of  mistakes  from  the  amount  completed

before  dividing.

E.

g->

Therefore,  whenever

the  time  limit  for  a  test  is  uniform,  the  rating  formula  is: