VARIABLES Subcommand (REGRESSION command)

VARIABLES names all the variables to be used in the analysis.

  • The minimum specification is a list of two variables or the keyword ALL or COLLECT. COLLECT, which must be specified in parentheses, is the default.
  • Only one VARIABLES subcommand is allowed, and it must precede any DEPENDENT or METHOD subcommands.
  • You can use keyword TO to refer to consecutive variables in the active dataset.
  • The order of variables in the correlation matrix constructed by REGRESSION is the same as their order on VARIABLES. If (COLLECT) is used, the order of variables in the correlation matrix is the order in which they are first listed on the DEPENDENT and METHOD subcommands.

ALL. Include all user-defined variables in the active dataset.

(COLLECT). Include all variables named on the DEPENDENT and METHOD subcommands. COLLECT is the default if the VARIABLES subcommand is omitted. COLLECT must be specified in parentheses. If COLLECT is used, the METHOD subcommands must specify variable lists.

Example

REGRESSION VARIABLES=(COLLECT)
 /DEPENDENT=SAVINGS
 /METHOD=STEP POP15 POP75 INCOME
 /METHOD=ENTER GROWTH.
  • COLLECT requests that the correlation matrix include SAVINGS, POP15, POP75, INCOME, and GROWTH. Since COLLECT is the default, the VARIABLES subcommand could have been omitted.
  • The DEPENDENT subcommand defines a single equation in which SAVINGS is the dependent variable.
  • The first METHOD subcommand requests that the block of variables POP15, POP75, and INCOME be considered for inclusion using a stepwise procedure.
  • The second METHOD subcommand adds variable GROWTH to the equation.