PRINT Subcommand (GENLIN command)

The PRINT subcommand is used to display optional output.

• If the PRINT subcommand is not specified, then the default output indicated below is displayed.
• If the PRINT subcommand is specified, then the GENLIN procedure displays output only for those keywords that are specified.

CORB. Correlation matrix for parameter estimates.

COVB. Covariance matrix for parameter estimates.

CPS. Case processing summary. For generalized estimating equations, this keyword also displays the Correlated Data Summary table. This is the default output if the PRINT subcommand is not specified.

DESCRIPTIVES. Descriptive statistics. Displays descriptive statistics and summary information about the dependent variable, covariates, factors. This is the default output if the PRINT subcommand is not specified.

FIT. Goodness of fit. For generalized linear models, displays deviance and scaled deviance, Pearson chi-square and scaled Pearson chi-square, log likelihood, Akaike's information criterion (AIC), finite sample corrected AIC (AICC), Bayesian information criterion (BIC), and consistent AIC (CAIC).

Note that when the scale parameter is fit using the deviance (/CRITERIA SCALE = DEVIANCE) or Pearson chi-square (/CRITERIA SCALE = PEARSON), the algorithm begins by assuming the scale parameter equals 1. Following estimation of the regression coefficients, the estimated scale parameter is calculated. Finally, estimated standard errors, Wald confidence intervals, and significance tests are adjusted based on the estimated scale parameter. However, in order to ensure fair comparison in the information criteria and the model fit omnibus test (see the SUMMARY keyword below), the log likelihood is not revised by the estimated scale parameter. Instead, when the scale parameter is fit using the deviance or Pearson chi-square, the log likelihood is computed with the scale parameter set equal to 1.

For generalized estimating equations, displays two extensions of AIC for model selection: Quasi-likelihood under the independence model criterion (QIC) for choosing the best correlation structure, and corrected quasi-likelihood under the independence model criterion (QICC) for choosing the best subset of predictors.

The quasi-likelihood functions are computed with the scale parameter set equal to a fixed value if a fixed value is specified on the /CRITERIA SCALE keyword. Otherwise, if /CRITERIA SCALE = MLE, DEVIANCE, or PEARSON, then the quasi-likelihood functions are computed with the scale parameter set equal to 1.

Goodness of fit statistics are not available for generalized estimating equations when the multinomial distribution is used. Thus, if the REPEATED subcommand and /MODEL DISTRIBUTION = MULTINOMIAL are specified, then the FIT keyword is silently ignored.

This is the default output if the PRINT subcommand is not specified.

GEF. General estimable function(s).

HISTORY (integer). Iteration history. For generalized linear models, displays the iteration history for the parameter estimates and log-likelihood, and prints the last evaluation of the gradient vector and the Hessian matrix. Also displays the iteration history for the profile likelihood confidence intervals (if requested via CRITERIA CITYPE = PROFILE) and for type I or III analyses (if requested via PRINT SUMMARY).

For generalized estimating equations, displays the iteration history for the parameter estimates, and prints the last evaluation of the generalized gradient and the Hessian matrix. Also displays the iteration history for type III analyses (if requested via PRINT SUMMARY).

The HISTORY keyword may be followed optionally by an integer n in parentheses, where the integer is greater than zero. The iteration history table displays parameter estimates for every n iterations beginning with the 0th iteration (the initial estimates). The default is to print every iteration (n = 1). If HISTORY is specified, then the last iteration is always displayed regardless of the value of n.

LAGRANGE. Lagrange multiplier test. For the normal, gamma, inverse Gaussian, and Tweedie distributions, displays Lagrange multiplier test statistics for assessing the validity of a scale parameter that is computed using the deviance or Pearson chi-square, or set at a fixed number. For the negative binomial distribution, tests the fixed ancillary parameter.

The LAGRANGE keyword is honored if MODEL DISTRIBUTION = NORMAL, GAMMA, IGAUSS, or TWEEDIE and CRITERIA SCALE = DEVIANCE, PEARSON, or number; or if MODEL DISTRIBUTION = NEGBIN(number) is specified. Otherwise the keyword is ignored and a warning is issued.

If the REPEATED subcommand is specified, then the LAGRANGE keyword is silently ignored.

LMATRIX. Set of contrast coefficient (L) matrices. Displays contrast coefficients for the default effects and for the estimated marginal means if requested.

MODELINFO. Model information. Displays the dataset name, dependent variable or events and trials variables, offset variable, scale weight variable, probability distribution, and link function.

For generalized estimating equations, also displays the subject variables, within-subject variables, and working correlation matrix structure.

This is the default output if the PRINT subcommand is not specified.

SOLUTION. Parameter estimates and corresponding statistics. This is the default output if the PRINT subcommand is not specified.

The SOLUTION keyword may be followed optionally by the keyword EXPONENTIATED in parentheses to display exponentiated parameter estimates in addition to the raw parameter estimates.

SUMMARY. Model summary statistics. Displays model fit tests, including likelihood ratio statistics for the model fit omnibus test, and statistics for the type I or III contrasts for each effect (depending on the CRITERIA ANALYSISTYPE specification). This is default output if the PRINT subcommand is not specified.

If the REPEATED subcommand is specified, then only the statistics for each effect are displayed.

WORKINGCORR. Working correlation matrix. This keyword is honored only if the REPEATED is in effect. Otherwise it is silently ignored.

NONE. No PRINT subcommand output. None of the PRINT subcommand output is displayed. If NONE is specified, then no other keywords are allowed on the PRINT subcommand.