StataNow 19.5: Feature Highlights
We’re excited to introduce the new features in StataNow 19.5.
1) Import data from Parquet files (November 12, 2025 update)
The new import parquet command reads Apache Parquet files and converts thm into a Stata dataset. Most Parquet data types and compression methods are supported. You can also import only a subset of columns or rows. This feature is especially useful for data engineers, data scientists, and analysts working with big data or cloud environments.
Learn more: https://www.stata.com/statanow/import-data-parquet-files
2) Causal mediation analysis with two mediators (November 12, 2025 update)
The mediate command now allows causal mediation analysis with two mediators. You can estimate natural indirect effects, natural direct effects, controlled direct effects, and mediator-specific effects. Both parallel and sequential mediation models are supported, and covariates can be added to outcome and mediator models. This update is relevant for researchers in epidemiology, public health, psychology, political science, economics, and sociology.
Learn more: https://www.stata.com/statanow/causal-mediation-analysis-two-mediators/
3) Execute selected lines in the Do-file Editor(September 25, 2025 update)
In the Do-file Editor, you can now run multiple highlighted lines directly. After execution, the cursor automatically moves to the next runnable line or block, skipping comments and blank lines.
Learn more: https://www.stata.com/new-in-stata/features/#selected_lines
4) Local backups for Do-files (September 25, 2025 update)
The Do-file Editor can now save backup files to an operating-system–specific directory on the local computer, instead of in the same folder as the original file. This option is available in the editor’s advanced settings.
Learn more: https://www.stata.com/new-in-stata/features/#backup_files
5) Enhanced grouping in graphs (September 25, 2025 update)
The commands graph bar, graph box, and graph dot now support the new assecondcategory option. This allows more flexible grouping: with multiple y-variables and one over() category, bars, boxes, or dots can be grouped in two steps—first by categories and then by y-variables.
Learn more: https://www.stata.com/new-in-stata/features/#grouping_enhancements
6) Power Analysis for Logistic Regression (May 21, 2025 update)
Design binary-outcome studies with confidence: power logistic computes required sample size, power, or effect size for a coefficient of interest, with tabular/graph outputs and support for additional nuisance covariates.
Learn more: https://www.stata.com/statanow/power-analysis-logistic-regression/
7) Local Average Treatment Effects (LATE) (August 13, 2025 update)
Estimate LATE for the population of compliers using lateffects. Choose among kappa-weighted and covariate-balancing estimators or IPWRA; built-in diagnostics help assess overlap and balance.
Learn more: https://www.stata.com/statanow/local-average-treatment-effects
8) More VCE Options for Linear Models (August 13, 2025 update)
Stata’s most used linear regression commands now support a broader set of variance–covariance (VCE) options:
- HAC and Driscoll–Kraay SEs,
- HC2/HC3 (with clustering),
- Hansen-style corrections,
- multiway clustering,
- unstructured time-covariances in panel GLS.
Learn more: https://www.stata.com/statanow/vce-options-for-linear-models
9) Do-file Editor Change History Ribbon (August 13, 2025 update)
The Do-file Editor can highlight what changed and what was reverted, helping you decide what to keep.
Video & info: https://www.stata.com/new-in-stata/do-file-editor/
Feature page: https://www.stata.com/new-in-stata/features/#history_ribbon
10) Improved Variable Name Truncation in the Data Editor (August 13, 2025 update)
Display long variable names using different truncation schemes to make variables easier to distinguish at a glance.
Learn more: https://www.stata.com/new-in-stata/features/#variable_truncation
Explore all StataNow additions: https://www.stata.com/new-in-stata/features/