Did participants in a training program obtain wages higher than their peers who did not participate? Treatment-effects estimators are used to measure the impact of an event, controlling for confounding factors such as age, gender, or level of education. Stata offers 20 built-in commands for treatment-effects estimation. In addition, many community-contributed commands exist to estimate treatment effects.
Ultimately, we need to communicate our results to others, and researchers typically do this by presenting tables of summary statistics and estimation results. Different disciplines and journals have their own styles, and an array of community-contributed commands for producing output exists to satisfy virtually all tastes. If Stata’s official abilities to produce output in Word, PDF, Excel, HTML, LaTeX, and other formats don’t fit your needs, a community-contributed command likely exists that does.
LIMITED DEPENDENT-VARIABLE MODELS
Not all dependent variables are continuous. Some are binary. Some are ordered. Some represent counts. Some are censored. Some are subject to sample selection. While Stata includes a spectrum of commands to handle such variables, the number of existing models is overwhelming and continues to grow. Fortunately, Stata’s built-in capability for programming maximum likelihood estimators makes implementing new models straightforward for user-programmers. Scores of community-contributed commands for limited dependent-variable models are now available for cross-sectional, panel, and multilevel datasets.
The focus of survival analysis is to model the amount of time required for an event to occur. Stata’s built-in survival analysis commands are widely recognized to be among the best in the industry, and practitioners have written additional commands to round out Stata’s offerings. Many community-contributed commands are available for cure and relative-risk models, discrete-time proportional-hazards models, and flexible parametric models.
When starting a research project, the data are almost never in the form you would like. Stata’s built-in data-management facilities are renowned, but you may come across a dataset that requires a custom level of manipulation beyond what you think Stata can do.
© Copyright 1996–2023 StataCorp LLC. All rights reserved.
Another Stata user has probably faced the same problem already and has made available a command to do the “heavy lifting”. Whether you need to convert data from a GIS program, manipulate value labels in your dataset, apply a linear filter, or create a complicated indicator variable, a community-contributed command is probably available to help.
MULTILEVEL AND CORRELATED DATA
Pupils are clustered within classrooms, which are clustered within schools, which are clustered within school districts. Consumers are clustered within neighborhoods, which are nested within towns, which are nested within metropolitan areas. Many datasets have observations that are nested within one or more larger groupings. Ignoring the correlations inherent in such data can result in inefficient or biased results. Stata offers 15 built-in commands for multilevel mixed-effects models. In addition, many community-contributed commands exist for multilevel data.
Econometricians frequently develop new estimators and tests, which are then implemented by Stata users. A variety of community-contributed commands are available for inequality measurement, interrupted time-series analysis models, demand-system estimation, and wage decompositions, to mention just a few areas of development.
Stata’s flexible graphics engine has motivated users to develop a variety of statistical graphs. Whether you need a specialized regression diagnostic plot to analyze the fit of your model, a plot of the cumulative distribution of a variable, a cycle plot to examine seasonality, a spine plot of two-way categorical data, a Bland–Altman plot to compare two assays, or a choropleth to map the spatial distribution of poverty, another Stata user has probably written the command you need.
Spatial data visualization. Univariate and multivariate statistical tests. The range of community-contributed features available is as diverse as the people who use Stata. Regardless of your field of study, there are community-contributed features that will complement your Stata experience.