COLLEGE STATION, Texas–(BUSINESS WIRE)–StataCorp, a leading developer of software for data science, announced the newest version of its software: Stata 18. Stata 18 is the latest in StataCorp’s almost 40-year history of providing powerful solutions for data manipulation, visualization, statistics, and automated reporting.
From designing a study to analyzing the collected data to then visualizing the results and, finally, efficiently communicating the study findings, the new Stata 18 provides a complete set of features for comprehensive and robust data analyses.
The latest version of the software builds on some of Stata’s core strengths, bringing in many new features to expand its already extensive data management and reporting features. These improvements range from all-new graph styles to a completely reimagined Data Editor, improving the way that researchers from any discipline work with their data and share their results.
“We designed Stata 18 to be intuitive for users so that it enhances their workflow,” said Chinh Nguyen, Vice President of Software Design. “It will expand their ability to gain insights and to share solutions, and ultimately to be more impactful with their research.”
Stata 18 introduces many new statistical features. The software offers broad analytical solutions for users from many disciplines, including economics, public policy, biostatistics, health research, business, and management. The new features will equip users with better tools for prediction, decision making, and modeling. For instance, school districts will be able to better measure student health outcomes. Clinicians will be able to stop a trial early if a treatment proves to be effective or ineffective. Policy makers will be able to detangle the effects of interest mediated through other factors.
“Our statistical developers conduct thorough research and acquire deep knowledge in broad scientific areas as they create the software,” said Yulia Marchenko, Vice President of Statistics and Data Science. “They know the software. They understand the theory behind the scientific methods they implement. This helps them anticipate and respond better to specific user needs, and design new features in a way that makes users’ analyses in Stata seamless, robust, and efficient.”
A wide range of statistical analyses in Stata 18, including Bayesian model averaging, causal mediation, heterogeneous difference-in-differences (DID), group sequential designs, and multilevel meta-analysis, to name a few, offer a powerful toolkit for many modeling needs.
For instance, Bayesian model averaging allows a researcher to account for the uncertainty when choosing a model. It provides a principled framework for borrowing information from multiple models, which are supported by the observed data, instead of relying on just one.
Causal mediation allows a researcher to disentangle the effects of interest when these effects are mediated through another factor, known as a mediator. Heterogeneous DID allows a researcher to model treatment effects that vary over time and over cohorts. Group sequential designs allow a researcher to stop a trial early for efficacy or futility. Finally, multilevel meta-analysis allows a researcher to combine the results of multiple studies while accounting for possible dependence among effect sizes that belong to the same level of hierarchy.
Other features new to Stata 18 include
- Tables of descriptive statistics
- Meta-analysis for prevalence
- Robust inference for linear models
- Wild cluster bootstrap
- Local projections for IRFs
- Model selection for ARIMA and ARFIMA
- Flexible demand systems
- TVCs with interval-censored Cox model
- Lasso for Cox model
- GOF plots for survival models
- IV quantile regression
- IV fractional probit model
- Corrected and consistent AICs
- New spline functions
- Alias variables across frames
- Frame sets
- Boost-based regular expressions
- Vectorized numerical integration
- New reporting features
- Do-file Editor enhancements
- Graph colors by variable
As with previous software releases, Stata 18 provides many features for users across disciplines. StataCorp remains focused on providing a software package that is broad in its reach but also meticulous in its execution. “Our users’ analysis and research needs are constantly evolving,” said Alan Riley, President. “Our goal is to deliver software that empowers our users to become better researchers who provide insight and make substantive contributions in their fields.”
For almost 40 years, StataCorp has been a leader in statistical and data science software. Stata software is a complete software package that provides solutions for all data science needs: data manipulation, visualization, statistics, and automated reporting.