Search
   >> Home >> Products >> Features >> Tests, predictions, and effects

Tests, predictions, and effects

Hypothesis testing

  • Wald test for linear constraints
  • Wald test (delta method) for nonlinear constraints
  • Likelihood-ratio test after any ML estimation
  • Bonferroni, Holm, and Šidák adjustments for multiple comparisons

Generalized testing

  • Ability to combine separate estimates into one combined estimate
  • Robust covariance matrix of combined estimates
  • Tests of linear and nonlinear combinations of estimates across fitted models
  • Point estimates and confidence intervals of linear and nonlinear combinations of estimates across fitted models

Predictions

  • Ability to obtain predicted values after all estimation commands
  • Predictor types that are tightly coupled to the estimation command
  • Default predicted value that is most relevant to the fitted model

Generalized predictions

  • Linear and nonlinear combinations of
    • Standard predictions
    • Equation index values
    • Estimated coefficients
    • Data
  • Inferential statistics for generalized predictions:
    • Point estimates
    • Standard errors
    • Variances
    • Wald test statistics
    • Significance levels
    • Pointwise confidence intervals

Postestimation statistics

  • Estimation sample summary statistics
  • Akaike and Bayesian information criteria
  • Covariance matrix analysis

Linear and nonlinear combinations of coefficients

  • Point estimates
  • Standard errors
  • Confidence intervals
  • Tests of significance
  • Covariances of transformations
  • Support for survey and clustered data

Effect sizes

  • Comparison of means
    • Cohen's d
    • Hedge's g
    • Glass's Δ
    • Point/biserial correlation
    • Confidence intervals
  • Variance explained by regression and ANOVA
    • Eta-squared—η2
    • Omega-squared—ω2
    • Confidence intervals

Marginal analysis

Watch Introduction to margins in Stata tutorials
Watch Profile plots and interaction plots in Stata tutorials

Forecast models

  • Combine results from multiple estimation commands
  • Specify identities and declare exogenous variables
  • Obtain dynamic and static forecasts
  • Use simulation methods to obtain predicition intervals
  • Specify alternative scenarios and perform "what-if" analyses

Hausman test

  • Test the independence of irrelevant alternatives (IIA) after
    • Multinomial logit
    • Conditional logistic regression
  • Test exogeneity or overidentifying restrictions for
    • Two-stage least squares (2SLS)
    • Three-stage least squares (3SLS)

Specification link test for single-equation models

    Save and restore estimation results

    • Save estimation results to disk
    • Compare models
    • Restore and perform predictions
    • Restore and perform tests

    Additional resources

    See contrasts and pairwise comparisons.

    See New in Stata 13 for more about what was added in Stata 13.

    The Stata Blog: Not Elsewhere Classified Find us on Facebook Follow us on Twitter LinkedIn Google+ Watch us on YouTube