Statistical Software Components Research Articles

Synth2: Synthetic control method with placebo tests, robustness test, and visualization

The synthetic control method (Abadie and Gardeazabal, 2003, American Economic Review 93: 113–132, Abadie, Diamond, and Hainmueller, 2010, Journal of the American Statistical Association 105: 493–505) is a popular method for causal inference in panel data with one treated unit that often uses placebo tests for statistical inference. While the synthetic control method can be implemented by the excellent command synth (Abadie, Diamond, and Hainmueller, 2011, Statistical Software Components S457334, Department of Economics, Boston College), it is still inconvenient for users to conduct placebo tests. As a wrapper program for synth, our proposed synth2 command provides convenient utilities to automate both in-space and in-time placebo tests, as well as the leave-one-out robustness test. Moreover, synth2 produces a complete set of graphs to visualize covariate or unit weights, covariate balance, actual or predicted outcomes, treatment effects, placebo tests, ratio of posttreatment mean squared prediction error to pretreatment mean squared prediction error, pointwise p-values (two-sided, right-sided, and left-sided), and the leave-one-out robustness test. We illustrate the use of the synth2 command by revisiting the classic example of California’s tobacco control program (Abadie, Diamond, and Hainmueller 2010).

Corrigendum

Statistical Analysis and Data Mining: The ASA Data Science JournalVolume 16, Issue 1 p. 92-92 CORRIGENDUMFree Access Corrigendum This article corrects the following: Neural network for univariate and multivariate nonlinearity tests Shapour Mohammadi, Volume 13Issue 1Statistical Analysis and Data Mining: The ASA Data Science Journal pages: 50-70 First Published online: November 18, 2019 First published: 21 November 2022 https://doi.org/10.1002/sam.11604AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinkedInRedditWechat In Mohammadi [1], The authors regret to inform that the citation for the bds.m code written by Professor Ludwig Kanzler were not included to Table 1. The updated Table 1 note is shown below. Note: Tests are Lee, White, and Granger (LWG); Terasvirta, Lee, and Granger (TLG); Tsay; Brock, Dechert, and Sheinkman (BDS); MC Leod Li (MCLoLi); Ramsy; Keenan; and AutoRegressive Conditional Heteroscedasticity (ARCH). Data generating process are white noise (WN), first-order autoregressive (AR1), second-order autoregressive (AR2), second order moving average (MA2), third order moving average (MA3), fifth order moving average (MA5), autoregressive moving average of various orders (ARMA1,1; ARMA2,1; ARMA2,2 and ARMA2,3). SizDist stands for distortion of empirical size from nominal. The BDS p-values are computed by using the code developed by Ludwig Kanzler. Reference: L. Kanzler, BDS: MATLAB module to calculate Brock, Dechert & Scheinkman test for independence, Statistical Software Components T871803, Boston College Department of Economics, Massachusetts, 1998. REFERENCE 1S. Mohammadi, Neural network for univariate and multivariate nonlinearity tests. Stat. Anal. Data Mining ASA Data Sci. J. 13 (2019), 9250– 9270. Volume16, Issue1February 2023Pages 92-92 ReferencesRelatedInformation

Binned scatterplots with marginal histograms: binscatterhist

I introduce binscatterhist, a command that extends the functionality of the popular binscatter command (Stepner, 2013, Statistical Software Components S457709, Department of Economics, Boston College). binscatter allows researchers to summarize the relationship between two variables in an informative and versatile way by collapsing scattered points into bins. However, information about the variables’ frequencies gets lost in the process. binscatterhist solves this issue by allowing the user to further enrich the graphs by plotting the variables’ underlying distribution. The binscatterhist command includes options for different regression methods, including reghdfe (Correia, 2014, Statistical Software Components S457874, Department of Economics, Boston College) and areg, and robust and clustered standard errors, with automatic reporting of estimation results and sample size.

Mobility data to aid assessment of human responses to extreme environmental conditions

Mobility data to aid assessment of human responses to extreme environmental conditions

Developing, maintaining, and hosting Stata statistical software on GitHub

The popularity of GitHub is growing, among not only software developers but also statisticians and data scientists. In this article, I discuss why social coding platforms such as GitHub are preferable for developing, documenting, maintaining, and collaborating on statistical software. Furthermore, I introduce the github command version 2.0 for Stata, which facilitates building, searching, installing, and managing statistical packages hosted on GitHub. I also provide a command for searching filenames in all Stata packages published on the Statistical Software Components Archive and GitHub to ensure unique filenames and package names, which is a common concern among Stata programmers. I make further suggestions to enhance the practice of developing and hosting statistical packages on GitHub as well as using them for data analysis.

Fast Poisson estimation with high-dimensional fixed effects

In this article, we present ppmlhdfe, a new command for estimation of (pseudo-)Poisson regression models with multiple high-dimensional fixed effects (HDFE). Estimation is implemented using a modified version of the iteratively reweighted least-squares algorithm that allows for fast estimation in the presence of HDFE. Because the code is built around the reghdfe package ( Correia, 2014 , Statistical Software Components S457874, Department of Economics, Boston College), it has similar syntax, supports many of the same functionalities, and benefits from reghdfe‘s fast convergence properties for computing high-dimensional leastsquares problems. Performance is further enhanced by some new techniques we introduce for accelerating HDFE iteratively reweighted least-squares estimation specifically. ppmlhdfe also implements a novel and more robust approach to check for the existence of (pseudo)maximum likelihood estimates.

On Identification and Estimation of Heckman Models

This article presents a Stata module to enable fixing the value of the correlation between the unobservables in Heckman models. These commands can solve two practical issues. First, for situations in which a valid exclusion restriction is not available, these commands enable exploring how the results could be affected by sample-selection bias. Second, stepping through values of this correlation can verify whether the global maximum of the likelihood function has been found. We provide several Stata commands to estimate these and related models with a fixed value of the correlation between the unobservables. These commands are available on the Boston College Statistical Software Components (SSC) archive and can be installed by typing ssc install fixedrho in Stata.

Advice on using heteroskedasticity-based identification

Lewbel (2012, Journal of Business and Economic Statistics 30: 67–80) provides a heteroskedasticity-based estimator for linear regression models containing an endogenous regressor when no external instruments or other such information is available. The estimator is implemented in the command ivreg2h by Baum and Schaffer (2012, Statistical Software Components S457555, Department of Economics, Boston College). In this article, we give advice and instructions to researchers who want to use this estimator.

When to Consult Precision-Recall Curves

ROC curves are commonly used to evaluate predictions of binary outcomes. When there are a small percentage of items of interest (as would be the case with fraud detection, for example), ROC curves can provide an inflated view of performance. This can cause challenges in trying to determine which set of predictions is better. This article discusses the conditions under which precision-recall curves may be preferable to ROC curves. As an illustrative example, we compare two commonly used fraud predictors (Beneish's (1999) M-score and Dechow et al.'s (2011) F-score) using both ROC and precision-recall curves. To aid the reader with using precision-recall curves, this article also introduces a Stata module to plot them. This module is now available on the Boston College Statistical Software Components (SSC) archive and can be installed by typing ssc install prtab in Stata.

Heckroc: ROC Curves for Selected Samples

ROC curves can be misleading when they are constructed with selected samples. In this article, we describe heckroc, a Stata command that implements a recently developed procedure for plotting ROC curves with selected samples. The command is available on the Boston College Statistical Software Components (SSC) and can be installed by typing SSC install heckroc in Stata. heckroc estimates the area under the ROC curve and a graphical display of the curve. A variety of plot options are available, including the ability to add confidence bands to the plot.

Introducing the StataStan Interface for Fast, Complex Bayesian Modeling Using Stan

In this article, we present StataStan, an interface that allows simulation-based Bayesian inference in Stata via calls to Stan, the flexible, open-source Bayesian inference engine. Stan is written in C++, and Stata users can use the commands stan and windowsmonitor to run Stan programs from within Stata. We provide a brief overview of Bayesian algorithms, details of the commands available from Statistical Software Components, considerations for users who are new to Stan, and a simple example. Stan uses a different algorithm than bayesmh, BUGS, JAGS, SAS, and MLwiN. This algorithm provides considerable improvements in efficiency and speed. In a companion article, we give an extended comparison of StataStan and bayesmh in the context of item response theory models.

Quantifying the Uptake of user-written Commands over Time

A major factor in the uptake of new statistical methods is the availability of user-friendly software implementations. One attractive feature of Stata is that users can write their own commands and release them to other users via Statistical Software Components at Boston College. Authors of statistical programs do not always get adequate credit, because programs are rarely cited properly. There is no obvious measure of a program's impact, but researchers are under increasing pressure to demonstrate the impact of their work to funders. In addition to encouraging proper citation of software, the number of downloads of a user-written package can be regarded as a measure of impact over time. In this article, we explain how such information can be accessed for any month from July 2007 and summarized using the new ssccount command.

Record Linkage Using Stata: Preprocessing, Linking, and Reviewing Utilities

In this article, we describe Stata utilities that facilitate probabilistic record linkage—the technique typically used for merging two datasets with no common record identifier. While the preprocessing tools are developed specifically for linking two company databases, the other tools can be used for many different types of linkage. Specifically, the stnd_compname and stnd_address commands parse and standardize company names and addresses to improve the match quality when linking. The reclink2 command is a generalized version of Blasnik's reclink (2010, Statistical Software Components S456876, Department of Economics, Boston College) that allows for many-to-one matching. Finally, clrevmatch is an interactive tool that allows the user to review matched results in an efficient and seamless manner. Rather than exporting results to another file format (for example, Excel), inputting clerical reviews, and importing back into Stata, one can use the clrevmatch tool to conduct all of these steps within Stata. This helps improve the speed and flexibility of matching, which often involves multiple runs.

Sensible Parameters for Univariate and Multivariate Splines

The package bspline, downloadable from Statistical Software Components, now has three commands. The first, bspline, generates a basis of Schoenberg B-splines. The second, frencurv, generates a basis of reference splines whose parameters in the regression model are simply values of the spline at reference points on the X axis. The recent addition, flexcurv, is an easy-to-use version of frencurv that generates reference splines with automatically generated, sensibly spaced knots. frencurv and flexcurv now have the additional option of generating an incomplete basis of reference splines, with the reference spline for a baseline reference point omitted or set to 0. This incomplete basis can be completed by adding the standard unit vector to the design matrix and can then be used to estimate differences between values of the spline at the remaining reference points and the value of the spline at the baseline reference point. Reference splines therefore model continuous factor variables as indicator variables (or “dummies”) model discrete factor variables. The method can be extended in a similar way to define factor-product bases, allowing the user to estimate factor-combination means, subset-specific effects, or even factor interactions involving multiple continuous or discrete factors.

From Resultssets to Resultstables in Stata

The listtab package supersedes the listtex package. It inputs a list of variables and outputs them as a table in one of several formats, including TEX, LATEX, HTML, Microsoft Rich Text Format, or possibly future XML-based formats. It works with a team of four other packages: sdecode, chardef, xrewide, and ingap, which are downloaded from the Statistical Software Components archive. The sdecode package is an extension of decode; it allows the user to add prefixes or suffixes and to output exponents as TEX, HTML, or Rich Text Format superscripts. It is called by another package, bmjcip, to convert estimates, confidence limits, and p-values from numeric to string. The chardef package is an extension of char define; it allows the user to define a characteristic for a list of variables, adding prefixes or suffixes. The xrewide package is an extension of reshape wide; it allows the user to store additional results in variable characteristics or local macros. The ingap package inserts gap observations into a dataset to form gap rows when the dataset is converted to a table. Together, these packages form a general toolkit to convert Stata datasets (or resultssets) to tables, complete with row labels and column labels. Examples are demonstrated using LATEX and also using the rtfutil package to create Rich Text Format documents. This article uses data from the Avon Longitudinal Study of Parents and Children, based at Bristol University, UK.

Comparing the Predictive Powers of Survival Models Using Harrell's C or Somers’ D

Medical researchers frequently make statements that one model predicts survival better than another, and they are frequently challenged to provide rigorous statistical justification for those statements. Stata provides the estat concordance command to calculate the rank parameters Harrell's C and Somers’ D as measures of the ordinal predictive power of a model. However, no confidence limits or p-values are provided to compare the predictive power of distinct models. The somersd package, downloadable from Statistical Software Components, can provide such confidence intervals, but they should not be taken seriously if they are calculated in the dataset in which the model was fit. Methods are demonstrated for fitting alternative models to a training set of data, and then measuring and comparing their predictive powers by using out-of-sample prediction and somersd in a test set to produce statistically sensible confidence intervals and p-values for the differences between the predictive powers of different models.

Density Distribution Sunflower Plots

Density distribution sunflower plots are used to display high-density bivariate data. They are useful for data where a conventional scatter plot is difficult to read due to overstriking of the plot symbol. The x-y plane is subdivided into a lattice of regular hexagonal bins of width w specified by the user. The user also specifies the values of l, d, and k that affect the plot as follows. Individual observations are plotted when there are less than l observations per bin as in a conventional scatter plot. Each bin with from l to d observations contains a light sunflower. Other bins contain a dark sunflower. In a light sunflower each petal represents one observation. In a dark sunflower, each petal represents k observations. (A dark sunflower with p petals represents between /2-pk k and /2+pk k observations.) The user can control the sizes and colors of the sunflowers. By selecting appropriate colors and sizes for the light and dark sunflowers, plots can be obtained that give both the overall sense of the data density distribution as well as the number of data points in any given region. The use of this graphic is illustrated with data from the Framingham Heart Study. A documented Stata program, called sunflower, is available to draw these graphs. It can be downloaded from the Statistical Software Components archive at http://ideas.repec.org/c/boc/bocode/s430201.html . (Journal of Statistical Software 2003; 8 (3): 1-5. Posted at http://www.jstatsoft.org/index.php?vol=8 .)

Review of Stata 7

Journal of Applied EconometricsVolume 16, Issue 5 p. 637-646 Software Review Review of Stata 7 Stanislav Kolenikov, Corresponding Author Stanislav Kolenikov skolenik@unc.edu Department of Statistics, University of North Carolina, Chapel Hill, NC 27599-3260, USA and Centre for Economic and Financial Research, Moscow, RussiaDepartment of Statistics, 117 New West, Cameron Avenue, University of North Carolina, Chapel Hill, NC, 27599-3260, USASearch for more papers by this author Stanislav Kolenikov, Corresponding Author Stanislav Kolenikov skolenik@unc.edu Department of Statistics, University of North Carolina, Chapel Hill, NC 27599-3260, USA and Centre for Economic and Financial Research, Moscow, RussiaDepartment of Statistics, 117 New West, Cameron Avenue, University of North Carolina, Chapel Hill, NC, 27599-3260, USASearch for more papers by this author First published: 22 October 2001 https://doi.org/10.1002/jae.631Citations: 4Read the full textAboutPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinked InRedditWechat Citing Literature Volume16, Issue5September/October 2001Pages 637-646 RelatedInformation