Stata 15 統計分析軟體

為什麼選擇 Stata ?
Stata是一個完整的集成軟件包,可提供您的所有數據科學需求 - 數據處理,可視化,統計和可重現的報告。
|
|

STATA
Stata is a complete, integrated statistical software package that provides everything you need for data analysis, data management, and graphics.
Bayesian analysis

- Thousands of built-in models
- Add your own models
- Adaptive Metropolis–Hastings
- Gibbs sampling
- Convergence diagnostics
- Posterior summaries
- Hypothesis testing
- Model comparison
IRT (item response theory)

- Binary response models—1PL, 2PL, 3PL
- Ordinal response models—graded response, partial credit, rating scale
- Nominal response model
- Hybrid models
- Item characteristic curves
- Test characteristic curves
- Item information function
Unicode

- Data
- Variable and value labels
- Variable names!
Integration with Excel

- Cell formatting
- Font formatting
- Insert Stata graphs
- Create cell formulas
Treatment effects

- Survival outcomes
- Endogenous treatments
- Balance diagnostics and tests
- Sampling weights
Multilevel survival models

- Random effects
- Crossed effects
- Two, three, higher level
- Right censoring
- Exponential, Weibull, ....
- Survey data
Multilevel models

- Survey data
- Multilevel sampling weights
- Survival models
- Denominator degrees of freedom
- Marginal predictions, means, effects
SEM (structural equation modeling)

- Satorra–Bentler adjustments
- Survival models
- Survey data
- Multilevel weights
- Marginal predictions, means, effects
Power and sample size

- Contingency tables
- Cochran–Mantel–Haenszel test
- Test for trend
- Matched case–control studies
- Survival analysis
Markov-switching models

- Autoregressive model
- Dynamic regression model
- State-dependent parameters
- Transition probabilities
- State membership probabilities
Survey statistics

- Multilevel models
- Survival models
- SEM (structural equation modeling)
- Multistage/multilevel weights
Panel-data survival models

- Random effects (intercepts)
- Random coefficients
- Right-censoring
- Exponential, Weibull, ...
- Survival graphs
Fractional outcome regression

- Fractions, proportions, ...
- Beta regression
- Probit and logit
- Heteroskedasticity
- Odds ratios
Marginal means and marginal effects

- Multiple outcomes
- Multiple equations
- Integrate over random effects
- Integrate over latent variables
More documentation

- Two all new manuals
- 12,000+ total pages
- Quick starts
- Thousands of worked examples
- Overview of statistical methods
More statistics

- Hurdle models
- Censored Poisson models
- Beta regression
- Structural break tests
- z tests comparing means
- More distribution functions
- Mersenne Twister