THE  TECHNIQUE  OF  GIVING  TESTS  163

quick  dash  like  this”—and  here  he  makes  a  quick,  straight
hne,  about  an  eighth  of  an  inch  long,  in  the  appropriate
place—“after  each  number  which  contains,”  etc.
The  instance  described  is  only  a  single,  small  detail
the  process  of  developing  directions  for  a  certain  test.
Every  set  of  directions  is  filled  with  minute  details  of
a  similar  nature.  In  fact,  it  is  practically  impossible  to
w °rk  out  a  set  of  directions  which  shall  be  perfect  at  the
°utset.  In  developing  the  directions  for  the  tests  which
yere  used  in  the  present  experiments  and  which  are  given,
ln  part,  in  the  Appendix,  it  was  the  practice  to  keep  a
blank  card  of  directions  for  each  test.  This  card  was
c arried  with  the  test  to  which  it  belonged,  and  whenever  a
Mistake  or  a  possible  ambiguity  in  the  wording  of  the
directions  was  noted,  that  fact  was  recorded  on  this  card.
this  manner,  through  a  long  series  of  tests  and  actual
experiences,  the  best  possible  set  of  directions  for  a  particular ­
  test  was  arrived  at.  It  is  safe  to  say  that  a  set  of
directions  for  any  one  test  can  not  be  made  perfect  or
fool  proof  until  it  has  been  tried  on  at  least  fifty  different
subjects.  No  one,  more  quickly  than  the  psychologist,
Earns  the  richness  of  meaning  which  even  the  simplest
' Vy ord  contains.  No  one,  therefore,  should  be  better  able
t°  guard  against  mistakes  arising  from  this  source.
To  minimize  the  mistakes  likely  to  arise  from  the
ambiguity  of  words,  the  psychologist  has  several  courses
°pen  to  him.  One  of  these  has  just  been  illustrated.  It
c °nsists  of  supplementing  his  words  with  the  action
"'hich  he  intends  the  subject  to  perform.  A  simple  demonoration
  like  this  is  often  more  lucid  than  a  paragraph  of
^utds.  However,  whenever  the  examiner  assumes  the
burden  of  demonstrating  his  own  directions,  care  must  be
taken  always  to  demonstrate  according  to  some  standard