New in Stata 12

Structural equation modeling (SEM)

  • Path diagrams
  • Graphical model builder
  • Standardized and unstandardized estimates
  • Modification indices
  • Direct and indirect effects
  • Score tests and Wald tests
  • Factors scores and other predictions
  • Goodness of fit
  • Estimation with groups and tests of invariance
  • Survey data and clustered data
  • Raw or statistical summary data
  • FIML estimation with missing at random (MAR) data
  • Maximum likelihood, ADF, and GMM estimation
  • Flexible extension of multivariate regression, instrumental variables, and simultaneous systems
  • Confirmatory factor analysis (CFA), correlated uniqueness models, latent growth models, multiple indicators and multiple causes (MIMIC), ...

Contrasts

  • Compare reference or adjacent categories
  • Compare to grand mean
  • Orthogonal polynomials
  • Treatment effects

Pairwise comparisons

  • Compare means, intercepts, or slopes
  • Compare odds ratios
  • Bonferroni, Scheffé, Tukey, Dunnett, and other adjustments

Margins plots

  • Profile and interaction plots
  • Margins, contrasts, and pairwise comparisons
  • Potential outcomes
  • Comparative graphs

Multiple imputation

  • Chained equations
  • Conditional imputation
  • Survey support
  • Impute separately within groups
  • Linear and nonlinear predictions
  • Measure simulation error
  • Panel data and multilevel models
  • Impute continuous, ordinal, cardinal, and count variables

ROC analysis

  • Parametric and nonparametric
  • Adjustments for covariates
  • Case-control regression models
  • Bootstrap and model-based SEs
  • Area under the curve (AUC) and partial AUC

Contour plots

Multilevel mixed-effects models

  • Complex survey data
  • Frequency and sampling weights
  • Robust and clustered SEs
  • Weighting at each level
  • Residual covariance structures: exponential, banded, and Toeplitz

Excel® import/export

  • Preview tool
  • Adjust import based on preview
  • More data management: ODBC connections strings, EBCDIC, rename groups of variables, ...

Unobserved components model (UCM)

  • Trend, seasonal, and cyclical components
  • Static and dynamic forecasts of components
  • Stochastic cycles

Automatic memory management

  • Automatically adjusts to dataset size
  • Tunable
  • Up to 1 terabyte of memory

ARFIMA

  • Long-memory processes
  • Fractional integration
  • Robust variance estimates
  • Static and dynamic forecasts
  • Linear constraints

Interface

  • Manage variables, storage types, notes, and formats without leaving the main interface
  • Select variables using filters
  • Filter prior commands and search results
  • Tabbed Viewer
  • Jump to dialogs, related commands, and sections in the online help
  • Hide, show, reorder, and filter variables in the Data Editor
  • Preview before pasting data
  • PDF export of results and graphs

Multivariate GARCH

  • Constant conditional correlations (CCC)
  • Dynamic conditional correlations (DCC)
  • Varying conditional correlations (VCC)
  • Multivariate normal and Students' t errors
  • Robust variance estimates
  • Level and variance predictions
  • Static and dynamic forecasts

Spectral density

  • Parametric estimates after ARIMA, ARFIMA, and UCM
  • Assess importance of frequencies

Installation Qualification

  • Downloadable tool
  • Report for submission to regulatory agencies

Time-series filters

  • Trend and cycle decompositions
  • Christiano-Fitzgerald band-pass filter
  • Baxter-King band-pass filter
  • Hodrick-Prescott high-pass filter
  • Butterworth high-pass filter

Stata/MP

  • More estimators
  • Up to 64 cores

Business calendars

  • Trading days
  • User definable
  • Lags and leads using business days
  • Conversions from standard calendar