Software | Gary King

PrivacyUnbiased

Fri, 01 Jan 2021 00:00:00 +0000

OpenDP: Developing Open Source Tools for Differential Privacy

Wed, 01 Jan 2020 00:00:00 +0000

QuickCode

Wed, 01 Jan 2020 00:00:00 +0000

Compactness: An R Package for Measuring Legislative District Compactness If You Only Know It When You See It

Mon, 01 Jan 2018 00:00:00 +0000

This software implements the methods in Kaufman, King, and Komisarchik, “How to Measure Legislative District Compactness If You Only Know It When You See It,” American Journal of Political Science. To deter gerrymandering, many U.S. state constitutions require legislative districts to be geographically “compact” (and a similar requirement holds explicitly or implicitly for numerous political jurisdictions around the world). Yet, the law offers few precise definitions other than “you know it when you see it,” which effectively implies a common understanding of the concept. In contrast, academics have shown that compactness has multiple dimensions and have generated many conflicting measures. The authors hypothesize that both are correct—that compactness is complex and multidimensional, but a single common understanding exists across people. They develop a survey to elicit this understanding, with high reliability (in data where the standard paired comparisons approach fails). They then create a statistical model that predicts, with high accuracy, solely from the geometric features of the district, compactness evaluations by judges and public officials responsible for redistricting, among many others. The project also offers compactness data from a validated measure for many state legislative and congressional districts, and software to compute this measure from any district.

PSI (Ψ): A Private Data Sharing Interface

Mon, 01 Jan 2018 00:00:00 +0000

Readme2: An R Package for Improved Automated Nonparametric Content Analysis for Social Science

Mon, 01 Jan 2018 00:00:00 +0000

An R package for estimating category proportions in an unlabeled set of documents given a labeled set, by implementing the method described in Jerzak, King, and Strezhnev (2023). This method is meant to improve on the ideas in Hopkins and King (2010), which introduced a quantification algorithm to estimate category proportions without directly classifying individual observations. This version of the software refines the original method by implementing a technique for selecting optimal textual features in order to minimize the error of the estimated category proportions. Automatic differentiation, stochastic gradient descent, and batch re-normalization are used to carry out the optimization. Other pre-processing functions are available, as well as an interface to the earlier version of the algorithm for comparison. The package also provides users with the ability to extract the generated features for use in other tasks.

Some scholars build models to classify documents into chosen categories. Others, especially social scientists who tend to focus on population characteristics, instead usually estimate the proportion of documents in each category—using either parametric “classify-and-count” methods or “direct” nonparametric estimation of proportions without individual classification. Unfortunately, classify-and-count methods can be highly model dependent or generate more bias in the proportions even as the percent of documents correctly classified increases. Direct estimation avoids these problems, but can suffer when the meaning of language changes between training and test sets or is too similar across categories. The underlying approach includes and optimizes continuous text features, along with a form of matching adapted from the causal inference literature.

Booc.Io: Software for an Education System With Hierarchical Concept Maps

Sun, 01 Jan 2017 00:00:00 +0000

Publication information:

Michail Schwab, Hendrik Strobelt, James Tompkin, Colin Fredericks, Connor Huff, Dana Higgins, Anton Strezhnev, Mayya Komisarchik, Gary King, and Hanspeter Pfister. 2017. "Booc.Io: An Education System With Hierarchical Concept Maps". IEEE Transactions on Visualization and Computer Graphics, 23, 1, Pp. 571-80.

Perusall

Thu, 01 Jan 2015 00:00:00 +0000

RobustSE

Thu, 01 Jan 2015 00:00:00 +0000

The RobustSE R package implements the generalized information matrix (GIM) test to detect model misspecification described in King and Roberts (2015). “Robust standard errors” are used in a vast array of scholarship to correct standard errors for model misspecification. However, when misspecification is bad enough to make classical and robust standard errors diverge, assuming that it is nevertheless not so bad as to bias everything else requires considerable optimism. And even if the optimism is warranted, settling for a misspecified model, with or without robust standard errors, will still bias estimators of all but a few quantities of interest. The accompanying article shows how to use robust standard errors as diagnostic tools via the GIM statistic (based on differences between robust and classical variance estimates), with practical illustrations via simulations and real examples. Open source software is available at https://github.com/IQSS/RobustSE and implements the test for linear, Poisson, and negative binomial regressions.

MatchingFrontier: R Package for Calculating the Balance-Sample Size Frontier

Wed, 01 Jan 2014 00:00:00 +0000

MatchingFrontier is an easy-to-use R Package for making optimal causal inferences from observational data. Despite their popularity, existing matching approaches leave researchers with two fundamental tensions. First, they are designed to maximize one metric (such as propensity score or Mahalanobis distance) but are judged against another for which they were not designed (such as L1 or differences in means). Second, they lack a principled solution to revealing the implicit bias-variance trade off: matching methods need to optimize with respect to both imbalance (between the treated and control groups) and the number of observations pruned, but existing approaches optimize with respect to only one; users then either ignore the other, or tweak it, usually suboptimally, by hand.

MatchingFrontier resolves both tensions by consolidating previous techniques into a single, optimal, and flexible approach. It calculates the matching solution with maximum balance for each possible sample size (N, N-1, N-2,…). It thus directly calculates the entire balance-sample size frontier, from which the user can easily choose one, several, or all subsamples from which to conduct their final analysis, given their own choice of imbalance metric and quantity of interest. MatchingFrontier solves the obvious joint optimization problem in one run, automatically, without manual tweaking, and without iteration. Although for each subset size k, there exist a huge (N choose k) number of unique subsets, MatchingFrontier includes specially designed fast algorithms that give the optimal answer, usually in a few minutes.

MatchingFrontier has officially been “Qualified for Scientific Use” by the U.S. Food and Drug Administration.

JudgeIt II: A Program for Evaluating Electoral Systems and Redistricting Plans

Fri, 01 Jan 2010 00:00:00 +0000

Authors: Andrew Gelman, Gary King, Andrew C. Thomas

JudgeIt allows a user to construct a model of a two-party election system over multiple election cycles, derive quantities of interest about the system through statistical estimation and simulation, and produce output summary statistics and graphical plots of those quantities. Some of the quantities of interest are based on partisan symmetry as a standard of fairness in legislative redistricting, such as partisan bias as the deviation from fairness and electoral responsiveness which indexes how party control of legislative seats responds to changes in a party's success at the polls even in a fair system. (A uniform consensus has existed in the academic literature since at least King and Browning (1987) on partisan symmetry as a standard for fairness, and even the U.S. Supreme Court now appears to agree; see Grofman and King (2007).) JudgeIt will also estimate and graph seats-votes curves, make specific vote and seat predictions for individual districts, and calculate numerous other relevant statistics.

The program can evaluate electoral systems (1) When an election already has taken place, (2) When an election has not been held yet but a new redistricting plan (or plans) has been proposed or implemented, and (3) When you wish to assess what an election would have been like if held under certain specified counterfactual conditions (such as if no minority districts had been drawn, or term limitations had prevented incumbents from running for reelection). The methods implemented in JudgeIt were developed in Gelman and King (1994).

Documentation: PDF
Installation: Load R, and type:
- install.packages("devtools")
  library(devtools)
  install_github("IQSS/JudgeIt")
Please send all questions, bugs, and requests via the JudgeIt GitHub issues page (legacy mailing list no longer linked from this site).

ReadMe: Software for Automated Content Analysis

Fri, 01 Jan 2010 00:00:00 +0000

This program will read and analyze a large set of text documents and report on the proportion of documents in each of a set of given categories.

Authors: Daniel Hopkins, Gary King, Matthew Knowles, Steven Melendez

The ReadMe software package for R takes as input a set of text documents (such as speeches, blog posts, newspaper articles, judicial opinions, movie reviews, etc.), a categorization scheme chosen by the user (e.g., ordered positive to negative sentiment ratings, unordered policy topics, or any other mutually exclusive and exhaustive set of categories), and a small subset of text documents hand classified into the given categories.

If used properly, ReadMe will report, normally within sampling error of the truth, the proportion of documents within each of the given categories among those not hand coded. ReadMe computes quantities of interest to the scientific community based on the distribution within categories but does so by skipping the more error prone intermediate step of classifing individual documents. Other procedures are also included to make processing text easy.

ReadMe implements methods described in Daniel Hopkins and Gary King, A Method of Automated Nonparametric Content Analysis for Social Science, American Journal of Political Science, 54, 1 (January 2010): 229--247. (Paper: Article | Abstract: HTML)

Related software Readme2 is available here.

Reporting Bugs and Issues: Please use our Github Issue form.
Questions and feature requests: Discuss the software on our Discussions page.
Documentation: readme.pdf explains how to install and use the package
ReadMe for R:
- To install the package make sure that you have the devtools package installed and then run in R:
  library(devtools)
  install_github("iqss-research/VA-package")
  install_github("iqss-research/ReadMeV1")
- Source code is available at: https://github.com/iqss-research/ReadMeV1
- Note: on Windows you will need to ensure that Python is installed before installing ReadMe. To install Python see: https://www.python.org/downloads/
License: Creative Commons Attribution- Noncommercial-No Derivative Works 3.0 License, for academic use only. A commerical (and industrial strength) version has been built by, licensed to, and offered by Crimson Hexagon.

AMELIA II: A Program for Missing Data

Thu, 01 Jan 2009 00:00:00 +0000

Authors: James Honaker, Gary King, Matthew Blackwell

Amelia II "multiply imputes" missing data in a single cross-section (such as a survey), from a time series (like variables collected for each year in a country), or from a time-series-cross-sectional data set (such as collected by years for each of several countries). Amelia II implements our bootstrapping-based algorithm that gives essentially the same answers as the standard IP or EMis approaches, is usually considerably faster than existing approaches and can handle many more variables. Unlike Amelia I and other statistically rigorous imputation software, it virtually never crashes (but please let us know if you find to the contrary!). The program also generalizes existing approaches by allowing for trends in time series across observations within a cross-sectional unit, as well as priors that allow experts to incorporate beliefs they have about the values of missing cells in their data. Amelia II also includes useful diagnostics of the fit of multiple imputation models. The program works from the R command line or via a graphical user interface that does not require users to know R.

Amelia is named after this famous missing person.

Multiple imputation involves imputing m values for each missing cell in your data matrix and creating m "completed" data sets. (Across these completed data sets, the observed values are the same, but the missing values are filled in with different imputations that reflect our uncertainty about the missing data.) After imputation, Amelia will then save the m data sets. You then apply whatever statistical method you would have used if there had been no missing values to each of the m data sets, and use a simple procedure to combine the results. Under normal circumstances, you only need to impute once and can then analyze the m imputed data sets as many times and for as many purposes as you wish. The advantage of Amelia is that it combines the comparative speed and ease-of-use of our algorithm with the power of multiple imputation, to let you focus on your substantive research questions rather than spending time developing complex application-specific models for nonresponse in each new data set. Unless the rate of missingness is exceptionally high, m=5 (the program default) will usually be adequate. Other methods of dealing with missing data, such as listwise deletion, mean substitution, or single imputation, are in common circumstances biased, inefficient, or both. When multiple imputation works properly, it fills in data in such a way as to not change any relationships in the data but which enables the inclusion of all the observed data in the partially missing rows.

Amelia II is a new program, and follows in the spirit with the same purpose as the first version of Amelia by James Honaker, Anne Joseph, Gary King, Kenneth Scheve, and Naunihal Singh.

Reporting Bugs and Issues: Please use our Github Issue form.
Questions and feature requests: Discuss the software on our Discussions page.
Github: https://github.com/IQSS/amelia
Documentation: PDF
AmeliaView for Windows (for those who don't know R): to install:
1. install the current version of R if you haven't already
2. download and run this file
3. click on the "AmeliaView" shortcut from the Desktop or the Start Menu.
Amelia for R: To install on any system: at the R command line, type
- install.packages("Amelia")
To use a development version of Amelia, enter the following commands at the R prompt:
- library(devtools)
  install_github("IQSS/Amelia")
To automatically combine multiply imputed data sets: in R see Zelig; In Stata see Clarify or Ken Scheve's MI program .
Papers related to Amelia:
- James Honaker and Gary King, "What to do About Missing Values in Time Series Cross-Section Data"American Journal of Political Science Vol. 54, No. 2 (April, 2010): Pp. 561-581. Article PDF
- Gary King, James Honaker, Anne Joseph, and Kenneth Scheve. "Analyzing Incomplete Political Science Data: An Alternative Algorithm for Multiple Imputation", American Political Science Review, Vol. 95, No. 1 (March, 2001): Pp. 49-69. Article
- Matthew Blackwell, James Honaker, and Gary King. A Unified Approach to Measurement Error and Missing Data: Overview and Details And Extensions both in Sociological Methods and Research, forthcoming.
A short course video circa 1999 which James, Ann, and Ken gave some years ago that explains mulitiple imputation in general, and the innovation in Amelia I in particular. Viewers will need to impute about 10 minutes of the video (at 10:19), which might have been when we reported the location of Ms. Earhart's plane.
A review of software for missing data

CEM: Coarsened Exact Matching Software

Thu, 01 Jan 2009 00:00:00 +0000

Authors: Stefano Iacus, Gary King, Giuseppe Porro

This program is designed to improve the estimation of causal effects via an extremely powerful method of matching that is widely applicable and exceptionally easy to understand and use (if you understand how to draw a histogram, you will understand this method). The program implements the Coarsened Exact Matching (CEM) algorithm described in:

"Causal Inference Without Balance Checking: Coarsened Exact Matching" (Political Analysis, 2012) and "Multivariate Matching Methods That are Monotonic Imbalance Bounding" (JASA, 2011), "CEM: Coarsened Exact Matching in Stata" (Stata Journal, 2009, with Matthew Blackwell), "CEM: Software for Coarsened Exact Matching." (Journal of Statistical Software, 2009), "A Theory of Statistical Inference for Matching Methods in Causal Research" (Political Analysis, 2019). See also An Explanation of CEM Weights.

Matching is a nonparametric method of preprocessing data to control for some or all of the potentially confounding influence of pretreatment control variables by reducing imbalance between the treated and control groups. After preprocessing in this way, any method of analysis that would have been used without matching can be applied to estimate causal effects, although some methods will have even better properties. CEM is a Monotonoic Imbalance Bounding (MIB) matching method --- which means that the balance between the treated and control groups is chosen by the user ex ante rather than discovered through the usual laborious process of checking after the fact and repeatedly reestimating, and so that adjusting the imbalance on one variable has no effect on the maximum imbalance of any other. CEM also strictly bounds through ex ante user choice both the degree of model dependence and the average treatment effect estimation error, eliminates the need for a separate procedure to restrict data to common empirical support, meets the congruence principle, is robust to measurement error, works well with multiple imputation methods for missing data, can be completely automated, and is extremely fast computationally even with very large data sets. After preprocessing data with CEM, the analyst may then use a simple difference in means or whatever statistical model they would have applied without matching. CEM also works well for multicategory treatments, determining blocks in experimental designs, and evaluating extreme counterfactuals.

CEM has officially been "Qualified for Scientific Use" by the U.S. Food and Drug Administration.

Reporting Bugs and Issues: Please use our Github Issue form.
Questions and feature requests: Discuss the software on our Discussions page.
CEM Package for R:
- Can be installed from CRAN: install.packages("cem")
- To install, from R:
  library(devtools); (install.packages("devtools") first if necessary)
  install_github("https://github.com/IQSS/cem.git")
- For documentation, from R, type library(cem), and then ?cem (or the published Journal of Statistical Software version)
- Github repository: https://github.com/IQSS/cem
CEM in MatchIt for R: Most of the features of CEM are also available through the R Package MatchIt: Nonparametric Preprocessing for Parametric Causal Inference.
CEM for SAS, by Stefano Verzillo, Paolo Berta, and Matteo Bossi
Download the SAS CEM Macro (Version: 2/2017, Questions: stefano.verzillo@ec.europa.eu)
See also JSCS article: "%CEM: A SAS macro to perform coarsened exact matching"
CEM for Stata (version 10 or later):
- To install, type:
  net from https://www.mattblackwell.org/files/stata
  net install cem
- You can also install from the SSC:
  ssc install cem
- For documentation, type help cem or download PDF (or the published version in The Stata Journal: PDF).
CEM for SPSS: Website
CEM for SQL (works with billions of observations): ZaliQL
CEM for Python: on github

OpenScholar

Thu, 01 Jan 2009 00:00:00 +0000

VA: Verbal Autopsies

Tue, 01 Jan 2008 00:00:00 +0000

VA is an easy-to-use R program that automates the analysis of verbal autopsy data. These data are widely used for estimating cause-specific mortality in areas without medical death certification.

Data on symptoms reported by caregivers along with the cause of death are collected from a medical facility, and the cause-of-death distribution is estimated in the population where only symptom data are available. Current approaches analyze only one cause at a time, involve assumptions judged difficult or impossible to satisfy, and require expensive, time consuming, or unreliable physician reviews, expert algorithms, or parametric statistical models. By generalizing current approaches to analyze multiple causes, King and Lu (2008) show how most of the difficult assumptions underlying existing methods can be dropped. These generalizations, which we implement here, also make physician review, expert algorithms, and parametric statistical assumptions unnecessary. While no method of analyzing verbal autopsy data can give accurate estimates in all circumstances, the procedure offered is conceptually simpler, less expensive, more general, as or more replicable, and easier to use in practice.

More generally, the software takes as input a multicategory variable D, and a set of dichotomous variables S (cause of Death and Symptoms, respectively, in verbal autopsy applications). Both variables exist in one data set (a hospital in the application) but only S exists in the population of interest. The goal of the procedure is to estimate the probability distribution (or histogram) of D in the population of interest.

For more information, see Gary King and Ying Lu, “Verbal Autopsy Methods with Multiple Causes of Death”.

Reporting bugs and issues: Please use our GitHub Issues form.
Questions and feature requests: Discuss the software on our Discussions page.
VA for R:
- To install: first, install and load the devtools library. Then, install_github("iqss-research/va-package").
- For built-in documentation in R: library(VA), and then ?va.
- GitHub repository: https://github.com/iqss-research/VA-package.
License: Creative Commons Attribution–Noncommercial–No Derivative Works 3.0 License, for academic use only. A commercial (and industrial-strength) version has been built by, and licensed to, Crimson Hexagon.

Anchors: Software for Anchoring Vignettes Data

Mon, 01 Jan 2007 00:00:00 +0000

Dataverse: Open Source Research Data Repository Software

Mon, 01 Jan 2007 00:00:00 +0000

MatchIt: Nonparametric Preprocessing for Parametric Causal Inference

Mon, 01 Jan 2007 00:00:00 +0000

Zelig: Everyone's Statistical Software

Sun, 01 Jan 2006 00:00:00 +0000

WhatIf: Software for Evaluating Counterfactuals

Sat, 01 Jan 2005 00:00:00 +0000

Authors: Heather Stoll, Gary King, Langche Zeng

Inferences about counterfactuals are essential for prediction, answering "what if" questions, and estimating causal effects. However, when the counterfactuals posed are too far from the data at hand, conclusions drawn from well-specified statistical analyses become based largely on speculation hidden in convenient modeling assumptions that few would be willing to defend.

What would happen if pigs could fly?The first known attempt to answer this question was in 1909 by J.T.C. Moore-Brabazon, who earlier the same year was the first British pilot to fly in Britain. On the left is Moore-Brabazon in his personal French-built Voisin aero plane. On the right is a pig in a wicker basket behind a sign that says "I am the first pig to fly."	Unfortunately, standard statistical approaches assume the veracity of the model rather than revealing the degree of model-dependence, which makes this problem hard to detect. WhatIf offers easy-to-apply methods to evaluate counterfactuals that do not require sensitivity testing over specified classes of models. If an analysis fails the tests offered here, then we know that substantive inferences will be sensitive to at least some modeling choices that are not based on empirical evidence, no matter what method of inference one chooses to use. WhatIf is also used to identify the areas of common support in causal inference. It is implemented in MatchIt and can easily process Zelig output objects so that counterfactuals can be evaluated, prior to computing quantities of interest, with only one additional command. WhatIf implements the methods for evaluating counterfactuals discussed in Gary King and Langche Zeng, 2006, "The Dangers of Extreme Counterfactuals," Political Analysis 14 (2): 131-159; and Gary King and Langche Zeng, 2007, "When Can History Be Our Guide? The Pitfalls of Counterfactual Inference," International Studies Quarterly 51 (March): 183-210.

What would happen if pigs could fly?The first known attempt to answer this question was in 1909 by J.T.C. Moore-Brabazon, who earlier the same year was the first British pilot to fly in Britain. On the left is Moore-Brabazon in his personal French-built Voisin aero plane. On the right is a pig in a wicker basket behind a sign that says "I am the first pig to fly."

Unfortunately, standard statistical approaches assume the veracity of the model rather than revealing the degree of model-dependence, which makes this problem hard to detect. WhatIf offers easy-to-apply methods to evaluate counterfactuals that do not require sensitivity testing over specified classes of models. If an analysis fails the tests offered here, then we know that substantive inferences will be sensitive to at least some modeling choices that are not based on empirical evidence, no matter what method of inference one chooses to use. WhatIf is also used to identify the areas of common support in causal inference. It is implemented in MatchIt and can easily process Zelig output objects so that counterfactuals can be evaluated, prior to computing quantities of interest, with only one additional command.

WhatIf implements the methods for evaluating counterfactuals discussed in Gary King and Langche Zeng, 2006, "The Dangers of Extreme Counterfactuals," Political Analysis 14 (2): 131-159; and Gary King and Langche Zeng, 2007, "When Can History Be Our Guide? The Pitfalls of Counterfactual Inference," International Studies Quarterly 51 (March): 183-210.

Reporting Bugs and Issues: Please use our Github Issue form.
Questions and feature requests: Discuss the software on our Discussions page.
WhatIf for R:
- Github: https://github.com/IQSS/WhatIf
- Installation directly from Github: devtools::install_github("IQSS/WhatIf")
- Documentation: Read on-line or Download PDF
Presentations on Whatif: Stoll: PPT, King: PDF

Gill/Murray/Cholesky/Factorization

Thu, 01 Jan 2004 00:00:00 +0000

Schnabel/Eskow/Cholesky/Factorization

Thu, 01 Jan 2004 00:00:00 +0000

YourCast

Thu, 01 Jan 2004 00:00:00 +0000

Authors: Federico Girosi, Gary King

YourCast is (open source and free) software that makes forecasts by running sets of linear regressions together in a variety of sophisticated ways. YourCast avoids the bias that results when stacking datasets from separate cross-sections and assuming constant parameters, and the inefficiency that results from running independent regressions in each cross-section.

The models enable you to have different covariates, or the same covariates with different meanings, in each cross-section. You may choose from a wide variety of smoothing techniques, such as assuming that the separate time series regressions in neighboring (or "similar") countries are alike, based on similarites in the coefficients (as in other approaches) or in the values or trends in the expected value of the dependent variable. This approach is advantageous because prior knowledge almost always exists about the dependent variable, and the expected value is always on the same metric even when including explanatory variables that differ in number or meaning in each country. You can also decide whether to smooth over indices that are geographic, grouped continuous variables (such as age groups), time, or interactions among these. For example, you can assume that, unless contradicted by the data, forecasts should be relatively smooth over time, or that the forecast time trends should be similar in adjacent age groups, or even that the differences in time trends between adjacent age groups stay roughly similar as they vary over countries.

The model works with time-series-cross-sectional data but also data for which the time series varies over more than one cross-section such as log-mortality over time by age, country, sex, and cause. The specific notion of "smoothness" or "similarity" used in YourCast is also your choice. The assumptions made by the statistical model are therefore governed your choices, and the sophistication of those assumptions and the degree to which they match empirical reality are, for the most part, limited only by what you may know or are willing to assume rather than arbitrary choices embedded in a mathematical model.

YourCast implements the methods introduced in Federico Girosi and Gary King's book manuscript on Demographic Forecasting, Princeton University Press, forthcoming. The present version is for those familar with R (if you're not, see Zelig); the next version will have more extensive preprocessing to make data input easier, and the version after that will have a GUI so that knowledge of R is unnecessary.

All questions, bugs and requests: use the IQSS file archive or R package documentation (legacy YourCast mailing list is no longer linked from this site).
Autocast for R:
- Documentation: Download PDF
- Link to Installation guide (see PDF documentation)
License: Creative Commons Attribution- Noncommercial-No Derivative Works 3.0 License, for academic use only.

Recommended Release

Version	Package		Date
1.6	Download (2.37 MB)	Release info	Sep 4 2013

Recent Releases

Version	Package		Date
1.5-2	Download (1.76 MB)	Release info	Aug 1 2013
1.5-1	Download (2.35 MB)	Release info	Apr 4 2012
1.1-12	Download (2.04 MB)	Release info	Feb 15 2011
1.1-11	Download (2.04 MB)	Release info	Sep 14 2010
1.1-10	Download (2.02 MB)	Release info	Mar 30 2010

CLARIFY: Software for Interpreting and Presenting Statistical Results

Wed, 01 Jan 2003 00:00:00 +0000

This is a set of easy-to-use tools that implement the techniques described in Gary King, Michael Tomz, and Jason Wittenberg's "Making the Most of Statistical Analyses: Improving Interpretation and Presentation". Winner of the Okidata Best Research Software Award from the American Political Science Association. These tools use Monte Carlo simulations to compute interpretable quantities from regression models and perform inference on them.

{clarify} for R

Implements predictions at representative values, average marginal effects, and any user-specified quantities of interest in a simulation framework, as well as visualization methods. {clarify} for R represents an evolution of the {Zelig} R package by restoring and adding to simulation-based functionality for translating hard-to-interpret coefficients into meaningful quantities of interest.
Authors: Noah Greifer, Steven Worthington, Stefano Iacus, and Gary King.
Website: https://iqss.github.io/clarify
GitHub: https://github.com/iqss/clarify
CRAN page: https://cran.r-project.org/package=clarify
See website for installation instructions, documentation, and examples.
Provides functionality previously provided by {Zelig}; see instructions on website for converting a {Zelig}-based workflow to one that uses {clarify} instead.

clarify for Stata

Implements predictions at representative values and visualization methods in a simulation framework.
Authors: Michael Tomz, Jason Wittenberg, and Gary King.
Github: https://github.com/iqss-research/clarify
Installation instructions and documentation are provided in a JSS Paper:
- Tomz, Michael, Jason Wittenberg, and Gary King. 2003. "Clarify: Software for Interpreting and Presenting Statistical Results." Journal of Statistical Software 8: 1–30. https://doi.org/10.18637/jss.v008.i01
A user donated wrapper from Fred Wolfe is available to automate clarify's simulation of dummy variables and can be installed with: ssc install qsim

EI: A Program for Ecological Inference

Wed, 01 Jan 2003 00:00:00 +0000

This program provides easy-to-use methods of running all the statistical procedures, diagnostics, and graphics developed in A Solution to the Ecological Inference Problem: Reconstructing Individual Behavior from Aggregate Data (Princeton University Press, 1997). The program has been rewritten from scratch in R: eiR on GitHub. The older Gauss version is still available here.

Reporting Bugs and Issues: Please use our Github Issue form.
Questions and feature requests: Discuss the software on our Discussions page.
EI for Gauss:
- To install, see the eiR repository on GitHub
- Documentation:PDF
- EzI version (standalone executable): Website.
EI for R: See the eiR repository on GitHub.

EzI: A(n Easy) Program for Ecological Inference

Wed, 01 Jan 2003 00:00:00 +0000

This software is no longer being actively updated. Previous versions and information about the software are archived here.

Authors: Kenneth Benoit, Gary King

This software is no longer being actively updated. Previous versions and information about the software are archived here.

This is a stand-alone, menu-oriented version of EI that runs under Windows. It does not require Gauss or any other software to run.

EzI does everything EI does and with fewer startup costs but, due to the lack of the Gauss command line, is somewhat less flexible. Winner of the APSA Research Software Award.

See readme.1st, included.
Github: https://github.com/iqss-research/ezi
All questions, bugs, requests: use the GitHub repository (legacy EI listserv is no longer linked from this site).

Recent Releases

Version	Package		Date
2.7:win	Download (2.12 MB)	Release info	Apr 14 2003

ReLogit: Rare Events Logistic Regression

Wed, 01 Jan 2003 00:00:00 +0000

Authors: Michael Tomz, Gary King, Langche Zeng

Both versions implement the suggestions described in Gary King and Langche Zeng's "Logistic Regression in Rare Events Data", "Explaining Rare Events in International Relations" and "Estimating Risk and Rate Levels, Ratios, and Differences in Case-Control Studies". Options for density case-control sampling designs are, at present, only available in the Gauss version.

Stata code is available at https://github.com/iqss-research/relogit.

Updated versions of ReLogit for R are available as part of the comprehensive statistical package Zelig: Everyone's Statistical Software. Zelig runs within R on all commonly used hardware and operating systems.
Reporting Bugs and Issues: Please use our Github Issue form.
Questions and feature requests: Discuss the software on our Discussions page.
Installing Relogit for Stata:
- Install from SSC by typing ssc install relogit.
- To install manually, download the package from the link below, and then put the files in the 'plus' directory, under 'r'. To find that directory, type sysdir.
- For documentation, type help relogit
- A helpful package to graph predictive probabilities and confidence intervals from relogitq is available here.

An Overview of the Virtual Data Center Project and Software

Mon, 01 Jan 2001 00:00:00 +0000

Software is now superseded by Dataverse.

Publication information

Micah Altman, Leonid Andreev, Mark Diggory, Gary King, Daniel L. Kiskis, Elizabeth Kolster, Michael Krot, and Sidney Verba. 2001. "An Overview of the Virtual Data Center Project and Software". JCDL '01: First Joint Conference on Digital Libraries, Pp. 203–4.

Abstract

In this paper, we present an overview of the Virtual Data Center (VDC) software, an open-source digital library system for the management and dissemination of distributed collections of quantitative data. (See Dataverse.) The VDC functionality provides everything necessary to maintain and disseminate an individual collection of research studies, including facilities for the storage, archiving, cataloging, translation, and on-line analysis of a particular collection. Moreover, the system provides extensive support for distributed and federated collections including: location-independent naming of objects, distributed authentication and access control, federated metadata harvesting, remote repository caching, and distributed "virtual" collections of remote objects.

For the current open-source research data repository software that continues this line of work, see Dataverse on this site.

MAXLIK

Thu, 01 Jan 1998 00:00:00 +0000

This software is no longer being actively updated. Previous versions and the information below are preserved for archival and teaching use.

MAXLIK is a set of Gauss programs and datasets (annotated for pedagogical purposes) to implement many of the maximum likelihood–based statistical models discussed in Gary King's book Unifying Political Methodology: The Likelihood Theory of Statistical Inference (University of Michigan Press, 1998), and used in Gary's courses. All datasets are real, not simulated.

Andrew Martin's related data sets in Stata: HTML (Washington University).

Recommended release

Version	Package	Date
1.0:dos-exe	DOS / Windows executable (archived distribution)	Jul 7, 1998

Original materials were distributed on Gary King's academic website; this page reproduces that documentation for use with modern environments where Gauss or legacy executables may no longer run.