Level of Significance

In testing a given  hypothesis  the maximum  probability with which  we would  be willing  to take  risk is called  level of  significance of the  test. This probability  often  denoted by  is generally specified before sample  are draw.

The level of significance  usually  employed in  testing of significance  are 0.05 ( or 5%) and 0.01 ( or 1%). If  for example  a 0.05 or 5% level of  significance  is chosen  in deriving a test of hypothesis  then there  are about  5 chances in 100 that we  would  reject the  hypothesis  when it  should  be accepted ( i .e.). we are about 95% confident that we made the right  decision. In  such a case we say that the hypothesis  has been  rejected at 5 % level  of significance which  means that we  could  be wrong  with  probability 0.05.

Note : Critical Region : A region  in the sample  space S which  amounts  to rejection of HO is termed  as critical region or region of  rejection.