Google
 

Trailing-Edge - PDP-10 Archives - decus_20tap2_198111 - decus/20-0026/ttest.doc
There are 2 other files named ttest.doc in the archive. Click here to see a list.
SUBROUTINE TTEST

PURPOSE
   TO FIND CERTAIN T-STATISTICS ON THE MEANS OF POPULATIONS.

USAGE
   CALL TTEST (A,NA,B,NB,NOP,NDF,ANS)

DESCRIPTION OF PARAMETERS
   A   - INPUT VECTOR OF LENGTH NA CONTAINING DATA.
   NA  - NUMBER OF OBSERVATIONS IN A.
   B   - INPUT VECTOR OF LENGTH NB CONTAINING DATA.
   NB  - NUMBER OF OBSERVATIONS IN B.
   NOP - OPTIONS FOR VARIOUS HYPOTHESES..
	 NOP=1--- THAT POPULATION MEAN OF B = GIVEN VALUE A.
		  (SET NA=1)
	 NOP=2--- THAT POPULATION MEAN OF B = POPULATION MEAN
		  OF A, GIVEN THAT THE VARIANCE OF B = THE
		  VARIANCE OF A.
	 NOP=3--- THAT POPULATION MEAN OF B = POPULATION MEAN
		  OF A, GIVEN THAT THE VARIANCE OF B IS NOT
		  EQUAL TO THE VARIANCE OF A.
	 NOP=4--- THAT POPULATION MEAN OF B = POPULATION MEAN
		  OF A, GIVEN NO INFORMATION ABOUT VARIANCES OF
		  A AND B.  (SET NA=NB)
   NDF - OUTPUT VARIABLE CONTAINING DEGREES OF FREEDOM ASSOCI-
	 ATED WITH T-STATISTIC CALCULATED.
   ANS - T-STATISTIC FOR GIVEN HYPOTHESIS.

REMARKS
   NA AND NB MUST BE GREATER THAN 1, EXCEPT THAT NA=1 IN
   OPTION 1. NA AND NB MUST BE THE SAME IN OPTION 4.
   IF NOP IS OTHER THAN 1, 2, 3 OR 4, DEGREES OF FREEDOM AND
   T-STATISTIC WILL NOT BE CALCULATED.	NDF AND ANS WILL BE
   SET TO ZERO.

SUBROUTINES AND FUNCTION SUBPROGRAMS REQUIRED
   NONE

METHOD
   REFER TO OSTLE, BERNARD, 'STATISTICS IN RESEARCH', IOWA
   STATE COLLEGE PRESS, 1954, CHAPTER 5.