Help System (web edition)

A sample mean may be compared to a known theoretical constant by the
usual t test, e.g.
      ANOVA IEADATA,30
      DEPENDENT=KNOWLDGE
      METHOD=TONESMPL(s)   CONSTANT=50
The default value for the constant is 0. Confidence interval for
the mean of the dependent variable is obtained if the desired
siginificance level is given as the option, e.g. METHOD=TONESMPL(0.05).

The equality of the means of two dependent variables may be tested
by the paired t test (2 dependent samples), e.g.

      DEPENDENT=KNOWLDGE,APPLICAT
      METHOD=TPAIRED(S)

where option S will cause the printout of the p-values corresponding to
the test statistics. If there are more than two variables then tests
are performed for all pairs of variables.

The t test for two dependent samples is a special caser of one-sample
repeated measurements, where each of the n subjects are measured at p
treatment levels. Usually, the independence of the errors at the p
levels of the same subject cannot be assumed. The p treatment levels
(mutually dependent var.) are given by the DEPENDENT parameter, e. g.
       DEPENDENT=KNOWLDGE,APPLICAT,MENTALPR,COMPREHN
The overall test that the means of the four variables are equal, is
performed automatically because there are more than two variables.

  G = Analysis of repeated mesurements when you have both 
      between and within subjects factors
  M = Multivariate analysis of (co)variance 
  A = More information on ANOVA operations