Analysing data from evaluations of educational interventions using a randomised controlled trial design. Various analytical tools to perform sensitivity analysis using different methods are supported (e.g. frequentist models with bootstrapping and permutations options, Bayesian models). The included commands can be used for simple randomised trials, cluster randomised trials and multisite trials. The methods can also be used more widely beyond education trials. This package can be used to evaluate other intervention designs using Frequentist and Bayesian multilevel models.
xtivdfreg implements the two-stage instrumental variables estimator for large-T panel data models, as developed by Norkute, Sarafidis, Yamagata, and Cui (J. Econometrics, 2021). The instruments are defactored to control for a multifactor error structure. Heterogeneous slope coefficients can be allowed using a mean-group estimator. The command accommodates unbalanced panel data and permits highly flexible instrumentation strategies.
yaxis maps an existing variable, meaning usually a numeric variable with integer codes and value labels, or equivalently a string variable, to a new variable with integer values 1 up and with value labels, sorted according to a specified criterion. The element arises from a leading application of the command. You have a categorical variable that would define an axis of a graph, or one dimension of a table (the rows, or the columns, say), but the existing order of categories is not ideal. Some graph and table commands offer sorting on the fly, but this command may help wherever other commands do not offer that.
ihelp displays the help documents of the specified commands via browsers. help displays the help documents in a simple but inconvenient manner, while Stata manual offers detailed command instructions, and the relevant PDF formats also have better reading experiences. ihelp can open more detailed help documents in PDF format via the browser, which can facilitate reading and translation; it can also list commands similar to the specified command from the Stata manual.
stackreg implements the stacked regression analysis which facilitates statistical testing in a multiple testing framework. The stacked regression approach was suggested e.g. by Weesie (STB, 1999) and Pei et al. (JBES, 2019). stackreg is closely realted to the Stata command suest. Unlike suest, which is extremely flexible in allowing inference involving regression models of different type, stackreg is confined to the linear model. However, in the context of the linear model, stackreg is more flexible than suest in several respects. In detail: (i) stackreg accomodates multi-way clustering (Cameron et al., JBES, 2011), if the community-distributed command cgmreg (by Jonah B. Gelbach and Douglas L. Miller) is installed; (ii) stackreg allows imposing cross-equation constraints by specifying the option constraints(); (iii) stackreg is a panel data command that accomodates fixed-effects estimation; (iv) unlike suest, stackrg applies a degrees-of-freedom adjustment that exactly reproduces the standard errors equation-by-equantion estimation yields; (v) stackreg allows for factor-variables in depvars. stackreg can also be regarded as more robust alternative to sureg and mvreg in particular. While these commands implement an FGLS estimator and, hence, rely on strong assumptions when estimating cross-equation coefficient covariances, stackreg uses clustering in the same fashion as suest. The stacked regression approach may alternatively implemented using the data management tool stack.
mlad maximizes a log-likelihood function where the likelihood function is programmed in Python. This enables the gradients and Hessian matrix to be obtained using automatic differentiation and to take advantage of using multiple CPUs. With large datasets mlad tends to be substantially faster than ml and has the important advantage that you don't have to derive the gradients and the Hessian matrix analytically as these are obtained using automatic differentiation.
flexmat creates flexible matrices / tables of text, numbers or both. flexmat hardly throws any comfortability error. So, we can add cells, rows, columns, or even previously stored matrices to the current matrix without worrying about comfortability or about mixing reals with string. flexmat is extremely flexible in adding and deleting columns, rows, or cells to an existing table. Tables can then be exported to Word, Excel or LaTeX using asdocx.
The command can easily transform country names in English to ISO 3-, 2-letter or numeric codes or build dimensions for a balanced panel dataset for the whole world (as of ISO's list of countries) in three lines of Stata code.
outdetect identifies outliers (extreme values) in the distribution of a variable, and assesses their impact on a selection of popular inequality and poverty measures. The procedure by which outliers are detected involves two steps. First, the distribution of the target variable is transformed to approach a standard normal distribution. Second, a threshold is applied to the transformed variable, to set the bounds of an outlier detection region (typically corresponding to the tails of the transformed distribution). Users are allowed to choose among a range of available transformations, or to let outdetect choose the best fitting transformation. The output of outdetect compares raw statistics (computed using the target variable as is), to trimmed” statistics (computed using just those observations that are not flagged as outliers), allowing the analyst to appreciate the impact of extreme values on final estimates. The sensitivity of estimates to outliers can also be gauged by producing diagnostic graphs.
conjoint can analyse and visualise conjoint (factorial) experiments. More specifically, conjoint can estimate average marginal component effects (AMCE) and marginal means (MM) following the methods described in Hainmueller et al., (2014) and Leeper et al, (2020) and implemented in the R packages, cjoint (Barari et al., 2018) and cregg (Leeper and Barnfield, 2020). conjoint can estimate these for fully randomised designs and AMCEs for designs with unlimited and complex profile constraints. conjoint can also calculate estimates across subgroups, with different baselevels (AMCEs only) and null hypothesis values (MMs only). The results can be simply and easily plotted via coefplot.
lxhuse provides easy access to a number of Stata-format datasets used in blogs of lianxh.cn. Those blogs can be easily searched with lianxh command in Stata Command Window. To use the auto_test.dta dataset, give the command lxhuse auto_test. If you receive an error message, check the web page listing these datasets. If a Stata data file has been saved in .zip format on the server (usually because it is very large), you may give the filename, including .zip, and the zip file will be copied to your working directory, unzipped, and read into Stata. If the dataset is declared as a timeseries or panel, the tsset command will be issued to display those characteristics. The ancilliary files can be downloaded using lxh get command.
The autofill program makes it easy to fill in missing values with existing values in the data. It applies to all types of missing values in the Stata and can operate on lots of variables simultaneously. In addition, it provides a way to fill the missing values by group.
This Gauss module estimates the dynamic asymmetric causality tests developed by Hatemi-J (2012, 2021). The subsamples can be determined by (a) fixed rolling window or (b) by recursive rolling window anchored to the start. The dynamic tests are estimated along with the bootstrap critical values with leverage adjustments for both 5% and 10% significance levels. It also estimates the dynamic symmetric causality tests developed by Hacker and Hatemi-J (2006, 2012). For technical details see (1) Hatemi-J A. (2021) Dynamic Asymmetric Causality Tests with an Application, Papers 2106.07612, arXiv.org. (2) Hatemi-J A. (2012) Asymmetric Causality Tests with an Application, Empirical Economics, 43, 447-456. (3) Hacker S. and Hatemi-J A. (2006) Tests for causality between integrated variables using asymptotic and bootstrap distributions: theory and application, Applied Economics, 38(13), 1489-1500. (4) Hacker S. and Hatemi-J A. (2012) A bootstrap test for causality with endogenous lag length choice: theory and application in finance, Journal of Economic Studies, 39(2), 144-160.
sblsf allows users to easily browse and access the posts of The Stata Forum in the stata window, including the title of the post, the number of visits and the number of replies
Nonparametric regressions are powerful statistical tools to model relationships between dependent and independent variables with minimal assumptions on the underlying functional forms. However, the added flexibility creates a curse of dimensionality, an induces a high computational cost in samples with even moderate sizes. As an alternative, semiparametric models that combine the flexibility of nonparametric regressions with the structure of standard models. VC_pack provides commands for the estimation of smooth varying-coefficient model (Hastie and Tibshirani, 1993, Journal of the Royal Statistical Society, Series B 55: 757–796), based on kernel regression methods, using a new set of commands within that aim to facilitate bandwidth selection and model estimation as well as create visualizations of the results.
xtbalance2 creates an indicator variable to identify a balanced subsample from an unbalanced dataset. The program tries to maximise the numbers of observations with respect to either the time dimension of the number of cross-sections/groups.
This package provides the egen function wmean(), mnemonic for weighted mean: it calculates (optionally) byable, (optionally) weighted, Arithmetic, Geometric or Harmonic mean.
The educationdata package makes it easy for you to download and format the school, school district, and college & university data from the Urban Institute Education Data Portal Application Programming Interface (API). For help troubleshooting, please visit https://github.com/UrbanInstitute/education-data-package-stata for commonly asked questions or to submit an issue.
odk2doc converts xlsForms programmed for survey tools such as ODK into easily readable docx-files. It offers a wide range of options to tailor the output to the user’s need to minimize any formatting effort. odk2doc facilitates the back-and-forth-conversion between the programmed questionnaire and the questionnaire discussed in the team or used for training or publication. odk2xls converts xlsForms programmed for survey tools such as ODK into easily readable xls/xlsx-files.
grpvars is a collection of commands to create and manage variable groups. A variable group is essentially a list of variables saved in a global macro with the name specified by the name() option. grpvars commands can conveniently manipulate these macros, and maintain their properties in memory. This is especially useful for managing sets of variables to be used in several tables. For example, you can specify the base model for the whole project, and rather than redefining the global macro for the variable list, add/remove/replace variables according to the specific needs of the current table. This enables one to more easily narrate the variable selection process. grpvars also maintains $refcat global macro if option title option is specified, to be used in estout.
