TEST Subcommand (NOMREG command)
The TEST
subcommand
allows you to customize your hypothesis tests by directly specifying
null hypotheses as linear combinations of parameters.
-
TEST
is offered only through syntax. - Multiple
TEST
subcommands are allowed. Each is handled independently. - The basic format for the
TEST
subcommand is an optional list of values enclosed in parentheses, an optional label in quotes, an effect name or the keywordALL
, and a list of values. - The value list preceding the first effect or the
keyword
ALL
are the constants to which the linear combinations are equated under the null hypotheses. If this value list is omitted, the constants are assumed to be all zeros. - The label is a string with a maximum length of 255 characters (or 127 double-byte characters). Only one label per linear combination can be specified.
- When
ALL
is specified, only a list of values can follow. The number of values must equal the number of parameters (including the redundant ones) in the model. - When effects are specified, only valid effects appearing
or implied on the
MODEL
subcommand can be named. The number of values following an effect name must equal the number of parameters (including the redundant ones) corresponding to that effect. For example, if the effect A*B takes up six parameters, then exactly six values must follow A*B. To specify the coefficient for the intercept, use the keywordINTERCEPT
. Only one value is expected to followINTERCEPT
. - When multiple linear combinations are specified within
the same
TEST
subcommand, use semicolons to separate each hypothesis. - The linear combinations are first tested separately for each logit and then simultaneously tested for all of the logits.
- A number can be specified as a fraction with a positive denominator. For example, 1/3 or –1/3 are valid, but 1/–3 is invalid.
- Effects appearing or implied on the
MODEL
subcommand but not specified on theTEST
are assumed to take the value 0 for all of their parameters.
Example
NOMREG
movie BY gender date
/CRITERIA = CIN(95) DELTA(0) MXITER(100) MXSTEP(5) LCONVERGE(0)
PCONVERGE(1.0E-6) SINGULAR(1.0E-8)
/INTERCEPT = EXCLUDE
/PRINT = CELLPROB CLASSTABLE FIT CORB COVB HISTORY(1) PARAMETER
SUMMARY LRT
/TEST (0 0) = ALL 1 0 0 0;
ALL 0 1 0 0 .
-
TEST
specifies two separate tests: one in which the coefficient corresponding to the first category for gender is tested for equality with zero, and one in which the coefficient corresponding to the second category for gender is tested for equality with zero.