»  Home »  Products »  Features »  Linear models

## Linear models

Linear regression

• Influence statistics and fit diagnostics
• Ramsey regression specification-error test for omitted variables
• Variance-inflation factors
• Cook’s distance
• COVRATIO
• DFBETAs
• DFITs
• Diagonal elements of hat matrix
• Residuals, standardized residuals, studentized residuals
• Standard errors of the forecast, prediction, and residuals
• Welsch distance
• Constraints
• Instrumental variables
• Difference-in-differences estimation New
• Heteroskedastic regression
• Truncated regression
• Errors in variables
• Multivariate regression
• Tests for heteroskedasticity
• Cook and Weisberg test
• Szroetzer’s rank test
• Information matrix test
• Cameron and Trivedi’s decomposition
• White’s test
• Tests for autocorrelation
• Durbin–Watson
• Durbin–Watson d statistic
• Breusch–Godfrey
• Tests for structural breaks
• ARCH LM test
• Moran's test for spatial dependence
• Diagnostic plots
• Component plus residual plot
• Leverage vs. squared residual plot
• Residual vs. fitted plot
• Residual vs. predictor
• Effect sizes
• Eta-squared—η 2
• Omega-squared—ω2
• Confidence intervals
• Fixed- and random-effects models for panel data
• Traditional, robust (Huber/White/sandwich), cluster–robust, bootstrap, or jackknife standard errors
• Robust regression
• Graph predictions and confidence intervals
• Newey–West estimator of variance
• Variance-weighted least squares
• GLM
• GLS for cross-sectional time-series data
• Multiple imputation
• Finite mixture models
• Bayesian estimation
Watch Fitting and interpreting regression models tutorials

Censored outcomes

• Interval censored (such as income reported in ranges)
• Tobit model
• Correlated data corrections to standard errors
• Heteroskedastic consistent standard errors
• Model heteroskedasticity
• Predictions
• Outcome in the absence of censoring
• Outcome conditional on being in the censoring interval
• Outcome with censoring imposed
• Probability of censoring
• Finite mixture models
• Bayesian estimation
• Interval regression with endogenous regressors, treatment effects, and sample selection

Sample-selection linear models

• Maximum likelihood and Heckman's two-step estimation
• Robust, cluster–robust, bootstrap, and jackknife standard errors
• Linear constraints
• Combine with endogenous regressors and treatment effects
• Linear and exponential
• Lower or upper boundary values
• Robust, cluster—robust, bootstrap, and jackknife standard errors

Stochastic frontier models

• Production and cost frontiers
• Half-normal, exponential, and truncated-normal distributions
• Modeling of conditional heteroskedasticity
• Median regression
• Regression of any quantile
• Interquantile range regression
• Standard errors
• Koenker and Bassett
• Robust — choose bandwidth and kernel
• Bootstrap
• Multiple imputation
• Support for a wide variety of models
• Component-plus-residual plots
• Support for zero-inflated regressors
• Combine endogeneity, Heckman-style selection, and treatment effects
• Linear regression
• Random effects in one or all equations
• Exogenous or endogenous regressors
• Exogenous or endogenous treatment assignment
• Binary treatment–untreated/treated
• Ordinal treatment levels–0 doses, 1 dose, 2 doses, etc.
• Endogenous selection using probit or tobit
• All standard postestimation command available, including predict and margins
• Multilevel random effects
• BLUP estimation
• Residual-error structures for linear models
• Standard errors of BLUPs
• Multiple imputation
• Bayesian estimation

Endogeneity and simultaneous systems

• Two-stage least-squares regression
• LIML estimation
• GMM estimation
• Instrumental variables
• Tests of instrumental relevance
• Tests of overidentifying restrictions
• Three-stage least-squares regression
• Linear constraints within and across equations
• Finite mixture models
• Linear regression with endogenous regressors, treatment effects, and sample selection

Seemingly unrelated regressions

• Linear constraints within and across equations
• View and run all postestimation features for your command
• Automatically updated as estimation commands are run
• Automatically create indicators based on categorical variables
• Form interactions among discrete and continuous variables Updated
• Include polynomial terms
• Perform contrasts of categories/levels
Watch Introduction to factor variables in Stata tutorials
• Estimated marginal means
• Marginal and partial effects
• Average marginal and partial effects
• Least-squares means
• Predictive margins
• Adjusted predictions, means, and effects
• Works with multiple outcomes simultaneously
• Contrasts of margins
• Pairwise comparisons of margins
• Profile plots
• Interaction plots
• Graphs of margins and marginal effects
Watch Introduction to margins in Stata tutorials
Watch Profile plots and interaction plots in Stata tutorials
• Analysis of main effects, simple effects, interaction effects, partial
• interaction effects, and nested effects
• Comparisons against reference groups, of adjacent levels, or against
• the grand mean
• Orthogonal polynomials
• Helmert contrasts
• Custom contrasts
• ANOVA-style tests
• Contrasts of nonlinear responses
• Balanced and unbalanced data
• Contrasts in odds-ratio metric
• Contrasts of means, intercepts, and slopes
• Graphs of contrasts
• Compare estimated means, intercepts, and slopes
• Compare marginal means, intercepts, and slopes
• Balanced and unbalanced data
• Nonlinear responses
• Multiple-comparison adjustments: Bonferroni, Šidák, Scheffé, Tukey HSD, Duncan, and Student–Newman–Keuls adjustments
• Group comparisons that are significant
• Graphs of pairwise comparisons

See tests, predictions, and effects.

See New in Stata 17 to learn about what was added in Stata 17.