Variable List (NOMREG command)

The variable list specifies the dependent variable and the factors in the model.

  • The dependent variable must be the first specification on NOMREG. It can be of any type (numeric or string). Values of the dependent variable are sorted according to the ORDER specification.

    ORDER = ASCENDING. Response categories are sorted in ascending order. The lowest value defines the first category, and the highest value defines the last category.

    ORDER = DATA. Response categories are not sorted. The first value encountered in the data defines the first category. The last distinct value defines the last category.

    ORDER = DESCENDING. Response categories are sorted in descending order. The highest value defines the first category, and the lowest value defines the last category.

  • By default, the last response category is used as the base (or reference) category. No model parameters are assigned to the base category. Use the BASE attribute to specify a custom base category.

    BASE = FIRST. The first category is the base category.

    BASE = LAST. The last category is the base category.

    BASE = value. The category with the specified value is the base category. Put the value inside a pair of quotes if either the value is formatted (such as date or currency) or if the dependent variable is the string type.

  • Factor variables can be of any type (numeric or string). The factors follow the dependent variable separated by the keyword BY.
  • Covariate variables must be numeric. The covariates follow the factors, separated by the keyword WITH.
  • Listwise deletion is used. If any variables in a case contain missing values, that case will be excluded.
  • If the WEIGHT command was specified, the actual weight values are used for the respective category combination. No rounding or truncation will be done. However, cases with negative and zero weight values are excluded from the analyses.

Example

NOMREG response (ORDER = DESCENDING BASE='No') BY factor1.
  • Values of the variable response are sorted in descending order, and the category whose value is No is the base category.

Example

NOMREG
  movie  BY gender date
  /CRITERIA = CIN(95) DELTA(0) MXITER(100) MXSTEP(5) LCONVERGE(0)
              PCONVERGE(0)
  /INTERCEPT = EXCLUDE
  /PRINT = CLASSTABLE FIT PARAMETER SUMMARY LRT .
  • The dependent variable is movie, and gender and date are factors.
  • CRITERIA specifies that the confidence level to use is 95, no delta value should be added to cells with observed zero frequency, and neither the log-likelihood nor parameter estimates convergence criteria should be used. This means that the procedure will stop when either 100 iterations or five step-halving operations have been performed.
  • INTERCEPT specifies that the intercept should be excluded from the model.
  • PRINT specifies that the classification table, goodness-of-fit statistics, parameter statistics, model summary, and likelihood-ratio tests should be displayed.