Gary King

Inducing Sustained Creativity and Diversity in Large Language Models

Sun, 01 Mar 2026 00:00:00 +0000

Abstract

We address a not-widely-recognized subset of exploratory search, where a user sets out on a typically long “search quest” for the perfect wedding dress, overlooked research topic, killer company idea, etc. The first few outputs of current large language models (LLMs) may be helpful but only as a start, since the quest requires learning the search space and evaluating many diverse and creative alternatives along the way. Although LLMs encode an impressive fraction of the world’s knowledge, common decoding methods are narrowly optimized for prompts with correct answers and thus return mostly homogeneous and conventional results. Other approaches, including those designed to increase diversity across a small set of answers, start to repeat themselves long before search quest users learn enough to make final choices, or offer a uniform type of “creativity” to every user asking similar questions. We develop a novel, easy-to-implement decoding scheme that induces sustained creativity and diversity in LLMs, producing as many conceptually unique results as desired, even without access to the inner workings of an LLM’s vector space. The algorithm unlocks an LLM’s vast knowledge, both orthodox and heterodox, well beyond modal decoding paths. With this approach, search quest users can more quickly explore the search space and find satisfying answers.

Citation: Queenie Luo, Gary King, Michael Puett, Michael D. Smith, 2026. “Inducing Sustained Creativity and Diversity in Large Language Models”

Article Links: Main Article, Supplementary Material

Attachments

Who's to Blame for Survey Instability: Respondents With Nonexistent Preferences or Researchers With Flawed Measures? (talk at Bocconi University, 3 24 2026)

Thu, 01 Jan 2026 00:00:00 +0000

Who's to Blame for Survey Instability: Respondents With Random Preferences or Researchers With Flawed Measures? (talk at Johns Hopkins University, 2 12 2026)

Thu, 01 Jan 2026 00:00:00 +0000

Assessing Differences in Country-Level Estimates of Maternal Mortality: A Comparison of GMatH, UN, and GBD Model Results for 2020

Wed, 01 Jan 2025 00:00:00 +0000

Correcting Measurement Error Bias in Conjoint Survey Experiments

Wed, 01 Jan 2025 00:00:00 +0000

Correcting Measurement Error Bias in Conjoint Survey Experiments (University of Central Florida)

Wed, 01 Jan 2025 00:00:00 +0000

Evaluating the Impacts of Swapping on the US Decennial Census

Wed, 01 Jan 2025 00:00:00 +0000

Experimental Evidence on the (Limited) Influence of Reputable Media Outlets

Wed, 01 Jan 2025 00:00:00 +0000

How American Politics Ensures Electoral Accountability in Congress

Wed, 01 Jan 2025 00:00:00 +0000

If a Statistical Model Predicts That Common Events Should Occur Only Once in 10,000 Elections, Maybe It's the Wrong Model

Wed, 01 Jan 2025 00:00:00 +0000

Interpersonal and Cross-Cultural Incomparability in Survey Research

Wed, 01 Jan 2025 00:00:00 +0000

Statistical Intuition Without Coding (or Teachers)

Wed, 01 Jan 2025 00:00:00 +0000

Statistically Valid Inferences from Differentially Private Data Releases, II: Extensions to Nonlinear Transformations

Wed, 01 Jan 2025 00:00:00 +0000

Statistically Valid Inferences from Privacy Protected Data (Pew Research Center)

Wed, 01 Jan 2025 00:00:00 +0000

Survey Estimates of Wartime Mortality

Wed, 01 Jan 2025 00:00:00 +0000

Automated Cognitive Debriefing

Mon, 01 Jan 2024 00:00:00 +0000

Correcting Measurement Error Bias in Conjoint Survey Experiments (Harvard Experiments Working Group)

Mon, 01 Jan 2024 00:00:00 +0000

Differentially Private Survey Research

Mon, 01 Jan 2024 00:00:00 +0000

Global Maternal Health Country Typologies: A Framework to Guide Policy

Mon, 01 Jan 2024 00:00:00 +0000

Global Maternal Mortality Projections by Urban Rural Locationand Education Level: A Simulation-Based Analysis

Mon, 01 Jan 2024 00:00:00 +0000

How American Politics Ensures Electoral Accountability in Congress (UCLA)

Mon, 01 Jan 2024 00:00:00 +0000

How to Measure Legislative District Compactness If You Only Know It When You See It (Harvard Law School)

Mon, 01 Jan 2024 00:00:00 +0000

Is Survey Instability Due to Respondents Who Don't Understand Politics or Researchers Who Don't Understand Respondents? (Caltech)

Mon, 01 Jan 2024 00:00:00 +0000

Reverse Engineering Chinese Government Information Controls (Harvard University)

Mon, 01 Jan 2024 00:00:00 +0000

Reverse Engineering Chinese Government Information Controls (SUNY New Paltz)

Mon, 01 Jan 2024 00:00:00 +0000

Statistically Valid Inferences from Privacy Protected Data (Stat188, Harvard University)

Mon, 01 Jan 2024 00:00:00 +0000

A Simulation-Based Comparative Effectiveness Analysis of Policies to Improve Global Maternal Health Outcomes

Sun, 01 Jan 2023 00:00:00 +0000

Correcting Measurement Error Bias in Conjoint Survey Experiments (Stanford University)

Sun, 01 Jan 2023 00:00:00 +0000

How American Politics Ensures Electoral Accountability in Congress (Center for American Political Studies, Harvard University)

Sun, 01 Jan 2023 00:00:00 +0000

How American Politics Ensures Electoral Accountability in Congress (Nuffield College, Oxford University)

Sun, 01 Jan 2023 00:00:00 +0000

How American Politics Ensures Electoral Accountability in Congress (Washington University in St. Louis)

Sun, 01 Jan 2023 00:00:00 +0000

How Science Works, and Some Advice (for You and for Me) (Department of Politics and International Relations, Oxford University)

Sun, 01 Jan 2023 00:00:00 +0000

Matching Methods for Observational and Experimental Causal Inference (Facultad Latinoamericana de Ciencias Sociales)

Sun, 01 Jan 2023 00:00:00 +0000

Simulation-Based Estimates and Projections of Global, Regional and Country-Level Maternal Mortality by Cause, 1990–2050

Sun, 01 Jan 2023 00:00:00 +0000

Solving Data Sharing Challenges Technologically through Differential Privacy (Wesleyan University)

Sun, 01 Jan 2023 00:00:00 +0000

Statistically Valid Inferences from Differentially Private Data Releases, With Application to the Facebook URLs Dataset

Sun, 01 Jan 2023 00:00:00 +0000

Statistically Valid Inferences from Privacy Protected Data

Sun, 01 Jan 2023 00:00:00 +0000

Statistically Valid Inferences from Privacy Protected Data (Washington University in St. Louis)

Sun, 01 Jan 2023 00:00:00 +0000

The Essential Role of Statistical Inference in Evaluating Electoral Systems: A Response to DeFord et Al.

Sun, 01 Jan 2023 00:00:00 +0000

An Improved Method of Automated Nonparametric Content Analysis for Social Science

Sat, 01 Jan 2022 00:00:00 +0000

Brief of Empirical Scholars As Amici Curiae in Support of Respondents

Sat, 01 Jan 2022 00:00:00 +0000

Noisy Data from the Noisy Census (Center for Discrete Mathematics and Theoretical Computer Science, Rutgers University)

Sat, 01 Jan 2022 00:00:00 +0000

Public Policy for the Poor? A Randomized Evaluation of the Mexican Universal Health Insurance Program (Harvard School of Public Health)

Sat, 01 Jan 2022 00:00:00 +0000

Rejoinder: Concluding Remarks on Scholarly Communications

Sat, 01 Jan 2022 00:00:00 +0000

Simplifying Matching Methods for Causal Inference (University of Wisconsin at Madison)

Sat, 01 Jan 2022 00:00:00 +0000

Statistically Valid Inferences from Privacy Protected Data (Deloitte)

Sat, 01 Jan 2022 00:00:00 +0000

Statistically Valid Inferences from Privacy Protected Data (Institute for Analytical Sociology, Norrköping, Sweden)

Sat, 01 Jan 2022 00:00:00 +0000

Statistically Valid Inferences from Privacy Protected Data (MIT Analytics Lab)

Sat, 01 Jan 2022 00:00:00 +0000

Statistically Valid Inferences from Privacy Protected Data (Princeton University)

Sat, 01 Jan 2022 00:00:00 +0000

Statistically Valid Inferences from Privacy Protected Data (SICSS, University of Rochester)

Sat, 01 Jan 2022 00:00:00 +0000

Statistically Valid Inferences from Privacy Protected Data (University of Wisconsin)

Sat, 01 Jan 2022 00:00:00 +0000

Video Presentations

Sat, 01 Jan 2022 00:00:00 +0000

A Theory of Statistical Inference for Ensuring the Robustness of Scientific Results

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

Cluster Analysis of Participant Responses for Test Generation or Teaching (2nd)

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

Designing Social Inquiry: Scientific Inference in Qualitative Research, New Edition

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

Education and Scholarship by Video

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

Empowering Social Science Research With Industry Partnerships (Dean's Symposium on Social Science Innovations, Harvard)

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

How to 'help Solve Society's Most Pressing Challenges' (quote from ALI Homepage) -- at the Harvard Advanced Leadership Initiative

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

How to Measure Legislative District Compactness If You Only Know It When You See It

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

How to Measure Legislative District Compactness If You Only Know It When You See It (Department of Biomedical Informatics, Harvard Medical School)

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

How to Measure Legislative District Compactness If You Only Know It When You See It (IPSA World Congress)

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

Letter to US Census Bureau: 'Request for Release of 'noisy Measurements File' by September 30 Along With Redistricting Data Products'

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

Participant Grouping for Enhanced Interactive Experience (4th)

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

Precision Mapping Child Undernutrition for Nearly 600,000 Inhabited Census Villages in India

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

PrivacyUnbiased

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

Scientific Inferences From Privatized Census Data (Federal Reserve Board, Panel)

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

Scientific Measurement in Redistricting Research (Princeton University)

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

Simplifying Matching Methods for Causal Inference (University of Wisconsin at Madison)

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

Survey Data and Human Computation for Improved Flu Tracking

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

There's a Simple Solution to the Latest Census Fight

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

UnbiasedPrivacy

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

Building an International Consortium for Tracking Coronavirus Health Status

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

Computational Social Science: Obstacles and Opportunities

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

Do Nonpartisan Programmatic Policies Have Partisan Electoral Effects? Evidence from Two Large Scale Experiments

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

Empowering Social Science to Understand and Ameliorate Major Challenges of Human Society (Federal Interagency Conference on Social Science and Big Data)

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

Evaluating COVID-19 Public Health Messaging in Italy: Self-Reported Compliance and Growing Mental Health Concerns

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

Expert Report of Gary King, in Bowyer et Al. V. Ducey (Governor) et Al., US District Court, District of Arizona

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

Instructional Support Platform for Interactive Learning Platforms (2nd)

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

OpenDP: Developing Open Source Tools for Differential Privacy

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

OpenDP: Developing Open Source Tools for Differential Privacy

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

Population-Scale Longitudinal Mapping of COVID-19 Symptoms, Behaviour and Testing

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

PrivacyUnbiased

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

QuickCode

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

QuickCode

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

Reverse Engineering Chinese Government Information Controls (Hebrew University of Jerusalem)

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

Simplifying Matching Methods for Causal Inference (Hebrew University of Jerusalem)

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

So You're a Grad Student Now? Maybe You Should Do This

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

Statistically Valid Inferences from Privacy Protected Data (Google, Inc)

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

Statistically Valid Inferences from Privacy Protected Data (Google)

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

Statistically Valid Inferences from Privacy Protected Data (Harvard University, Applied Statistics Workshop)

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

Statistically Valid Inferences from Privacy Protected Data (Harvard, Privacy Tools Project)

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

Statistically Valid Inferences from Privacy Protected Data (Interagency Arctic Research Policy Committee)

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

Statistically Valid Inferences from Privacy Protected Data (Webcast, Project TIER)

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

The 'Math Prefresher' and The Collective Future of Political Science Graduate Training

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

The SilverLining Project: Finding Social Good in Clouds on the Dark Web

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

Theoretical Foundations and Empirical Evaluations of Partisan Fairness in District-Based Democracies

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

A New Model for Industry-Academic Partnerships

Tue, 01 Jan 2019 00:00:00 +0000

A Theory of Statistical Inference for Matching Methods in Causal Research

Tue, 01 Jan 2019 00:00:00 +0000

Cluster Analysis of Participant Responses for Test Generation or Teaching

Tue, 01 Jan 2019 00:00:00 +0000

Ecological Regression With Partial Identification

Tue, 01 Jan 2019 00:00:00 +0000

How the News Media Activate Public Expression and Influence National Agendas

Tue, 01 Jan 2019 00:00:00 +0000

How the News Media Activate Public Expression and Influence National Agendas (University of Minho)

Tue, 01 Jan 2019 00:00:00 +0000

How to Measure Legislative District Compactness If You Only Know It When You See It

Tue, 01 Jan 2019 00:00:00 +0000

How to Measure Legislative District Compactness If You Only Know It When You See It

Tue, 01 Jan 2019 00:00:00 +0000

How to Measure Legislative District Compactness If You Only Know It When You See It

Tue, 01 Jan 2019 00:00:00 +0000

How to Measure Legislative District Compactness If You Only Know It When You See It

Tue, 01 Jan 2019 00:00:00 +0000

How to Measure Legislative District Compactness If You Only Know It When You See It (University of Chicago)

Tue, 01 Jan 2019 00:00:00 +0000

Instructional Support Platform for Interactive Learning Platforms

Tue, 01 Jan 2019 00:00:00 +0000

Participant Grouping for Enhanced Interactive Experience (3rd)

Tue, 01 Jan 2019 00:00:00 +0000

Simplifying Matching Methods for Causal Inference

Tue, 01 Jan 2019 00:00:00 +0000

Simplifying Matching Methods for Causal Inference (University of Minho)

Tue, 01 Jan 2019 00:00:00 +0000

Statistically Valid Inferences from Privacy Protected Data (Microsoft)

Tue, 01 Jan 2019 00:00:00 +0000

Statistically Valid Inferences from Privacy Protected Data (University of Chicago)

Tue, 01 Jan 2019 00:00:00 +0000

Stimulating Online Discussion in Interactive Learning Environments

Tue, 01 Jan 2019 00:00:00 +0000

Systems and Methods for Keyword Determination and Document Classification from Unstructured Text

Tue, 01 Jan 2019 00:00:00 +0000

Why Propensity Scores Should Not Be Used for Matching

Tue, 01 Jan 2019 00:00:00 +0000

Big Data Is Not About the Data!

Mon, 01 Jan 2018 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

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.

Edited Transcript of a Talk on Partisan Symmetry at the 'Redistricting and Representation Forum'

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

How the News Media Activate Public Expression and Influence National Agendas

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

How the News Media Activate Public Expression and Influence National Agendas

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

How the News Media Activate Public Expression and Influence National Agendas

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

How the News Media Activate Public Expression and Influence National Agendas

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

How the News Media Activate Public Expression and Influence National Agendas

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

How the News Media Activate Public Expression and Influence National Agendas

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

How to Measure Legislative District Compactness If You Only Know It When You See It

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

Management of Off-Task Time in a Participatory Environment

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

Matching Methods for Causal Inference

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

PSI (Ψ): A Private Data Sharing Interface

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

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

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.

Reverse Engineering Chinese Government Information Controls

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

Reverse Engineering Chinese Government Information Controls

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

Sequential Experiments and News Media Effects

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

Simplifying Matching Methods for Causal Inference

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

Simplifying Matching Methods for Causal Inference

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

Simplifying Matching Methods for Causal Inference

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

Simplifying Matching Methods for Causal Inference

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

Use of a Social Annotation Platform for Pre-Class Reading Assignments in a Flipped Introductory Physics Class

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

Suggestions for Changes in Journal Publication Rules

Fri, 24 Mar 2017 12:00:00 +0000

(This was originally a post on the now defunkt Perestroika mailing list, on 9/27/10.)

I have two suggestions stemming from the discussion over the last few weeks.

Well before the Perestroika list started, many people have expressed complaints about how the American Political Science Review (APSR) and some other journals poorly represent the work of some; despite some changes, the complaints haven't subsided a lot. Since the APSR treats journal space as a scarce resource, it should not be a surprise to all of us political scientists that we still see lots of political discussions like these surrounding the allocation of those scarce resources. However, more recently, the world seems to have passed a threshold in publication where online is as good or better than print. If you don't already find it more convenient to look for an article in jstor sitting at your desk than reaching 'all the way' behind you to grab the print version, you will soon. Plus it's much easier to search electronic versions, and a vast amount of value-added information is being created with the digital but not print versions – comments, collaborative highlighting & note taking, social media posts right from the publications, etc., etc. Moreover, in many areas of scholarship, if you can't find prior research through Google or Bing, it just doesn't exist.

Whether this change is good or bad is a good question but not my point. Instead, I suggest we ask the APSR and APSA to recognize this change and respond to it, since when scarce resources become plentiful, many problems are automatically solved. And acting as if they are still scarce only perpetuates unnecessary division. So instead of pushing the APSR to publish more works like whatever we each do, why not push them to vastly increase the number of articles published? The marginal cost of publishing more articles is now nearly zero. My own view is that the threshold for publication should be something simple like whether the article represents a positive contribution to our knowledge or understanding of the world (or to something!); if yes, then publish. If its wrong, or misleading, or unclear, or dumb, or fraudulent, then reject. And if reviewers can "revise and resubmit" the author into doing a better job, then great. But we don't need reviewers and editors deciding on assessments of "importance", "area", "quant vs qual balance", or other irrelevant, or essentially political matters.

The Internet (and searches that return >2M items, but ranked so that the one you want is first) is ample proof that more information doesn't hurt anyone. When the press was actually a physical press, publication was expensive and the presses became the gatekeepers to their pocketbooks (and the visibility of our work); now publication is almost free. If someone wants to have a series of awards for the best articles, or articles that are above some higher threshold, or which meet some criteria such as area or balance or anything else, then fine. Let the politics continue around these awards, rather than what it does today, which is to essentially ensure that some types of works or some works do not see the light of day.

I'll go another step. The point of the APSA is the creation, dissemination, and preservation of knowledge about political science (and a vast array of supporting activities). To achieve these goals better, why not make the APSR open source and free? Open source journals have more readers (especially in the developing world) and a bigger impact on the rest of the scholarly literature. The APSA as an association will still do very well financially, and its mission will be achieved at a much higher level.

So how about it? Encourage the APSR to publish more – without discriminating at all based on area and type of work and only on quality – and to make the proceeds of our work available for free to the world.

A Unified Approach to Measurement Error and Missing Data: Details and Extensions

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

A Unified Approach to Measurement Error and Missing Data: Overview and Applications

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

An Improved Method of Automated Nonparametric Content Analysis for Social Science

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

Analyzing Social Media Data in China

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

Big Data Is Not About the Data!

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

Big Data Is Not About the Data!

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

Big Data Is Not About the Data!

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

Big Data Reveals Made Up Data: How the Chinese Government Fabricates Social Media Posts for Strategic Distraction, Not Engaged Argument

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

Booc.Io: An Education System With Hierarchical Concept Maps

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

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.

Brief of Heather K. Gerken, Jonathan N. Katz, Gary King, Larry J. Sabato, and Samuel S.-H. Wang As Amici Curiae in Support of Appellees

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

Computer-Assisted Keyword and Document Set Discovery from Unstructured Text

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

Detecting and Reducing Model Dependence in Causal Inference

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

Fabricating News In Chinese Social Media

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

How the Chinese Government Fabricates Social Media Posts for Strategic Distraction, Not Engaged Argument

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

How the Chinese Government Fabricates Social Media Posts for Strategic Distraction, Not Engaged Argument

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

How the Chinese Government Fabricates Social Media Posts for Strategic Distraction, Not Engaged Argument

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

How the Chinese Government Fabricates Social Media Posts for Strategic Distraction, Not Engaged Argument

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

How the Chinese Government Fabricates Social Media Posts for Strategic Distraction, Not Engaged Argument

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

How the Chinese Government Fabricates Social Media Posts for Strategic Distraction, Not Engaged Argument

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

How the News Media Activate Public Expression and Influence National Agendas

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

How to Conquer Partisan Gerrymandering

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

How to Measure Legislative District Compactness If You Only Know It When You See It

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

How to Measure Legislative District Compactness If You Only Know It When You See It

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

Matching Methods for Causal Inference and 21 Other Topics

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

OpenScholar

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

Reverse Engineering Chinese Government Information Controls

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

Reverse Engineering Chinese Government Information Controls

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

Simplifying Matching Methods for Causal Inference

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

Simplifying Matching Methods for Causal Inference

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

Simplifying Matching Methods for Causal Inference

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

Simplifying Matching Methods for Causal Inference

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

The Balance-Sample Size Frontier in Matching Methods for Causal Inference

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

Thresher (acquired by Two Six Technologies, a Carlyle Group Company)

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

An Improved Method of Automated Nonparametric Content Analysis for Social Science

Fri, 01 Jan 2016 00:00:00 +0000

Big Data Is Not About the Data!

Fri, 01 Jan 2016 00:00:00 +0000

Big Data Is Not About the Data!

Fri, 01 Jan 2016 00:00:00 +0000

Big Data Is Not About the Data!

Fri, 01 Jan 2016 00:00:00 +0000

Big Data Is Not About the Data!

Fri, 01 Jan 2016 00:00:00 +0000

Big Data Is Not About the Data!

Fri, 01 Jan 2016 00:00:00 +0000

Big Data Is Not About the Data! The Power of Modern Analytics

Fri, 01 Jan 2016 00:00:00 +0000

Comment on 'Estimating the Reproducibility of Psychological Science'

Fri, 01 Jan 2016 00:00:00 +0000

Cross-Classroom and Cross-Institution Item Validation

Fri, 01 Jan 2016 00:00:00 +0000

Discovering and Explaining Systematic Bias and Nontransparency in US Social Security Administration Forecasts

Fri, 01 Jan 2016 00:00:00 +0000

Effectiveness of the WHO Safe Childbirth Checklist Program in Reducing Severe Maternal, Fetal, and Newborn Harm: Study Protocol for a Matched-Pair, Cluster Randomized Controlled Trial in Uttar Pradesh, India

Fri, 01 Jan 2016 00:00:00 +0000

Explaining Systematic Bias and Nontransparency in US Social Security Administration Forecasts

Fri, 01 Jan 2016 00:00:00 +0000

How Human Subjects Research Rules Mislead You and Your University, and What to Do About It

Fri, 01 Jan 2016 00:00:00 +0000

How the Chinese Government Fabricates Social Media Posts for Strategic Distraction, Not Engaged Argument

Fri, 01 Jan 2016 00:00:00 +0000

How the Chinese Government Fabricates Social Media Posts for Strategic Distraction, Not Engaged Argument

Fri, 01 Jan 2016 00:00:00 +0000

How the Chinese Government Fabricates Social Media Posts for Strategic Distraction, Not Engaged Argument

Fri, 01 Jan 2016 00:00:00 +0000

Introduction to Perusall

Fri, 01 Jan 2016 00:00:00 +0000

Method and Apparatus for Selecting Clusterings to Classify a Data Set

Fri, 01 Jan 2016 00:00:00 +0000

Preface: Big Data Is Not About the Data!

Fri, 01 Jan 2016 00:00:00 +0000

Reverse-Engineering Censorship in China

Fri, 01 Jan 2016 00:00:00 +0000

Scoring Social Security Proposals: Response from Kashin, King, and Soneji

Fri, 01 Jan 2016 00:00:00 +0000

Simplifying Matching Methods for Causal Inference

Fri, 01 Jan 2016 00:00:00 +0000

Systems and Methods for Calculating Category Proportions

Fri, 01 Jan 2016 00:00:00 +0000

The C-SPAN Archives As The Policymaking Record of American Representative Democracy: A Foreword

Fri, 01 Jan 2016 00:00:00 +0000

The Next Big [Social Science] Thing

Fri, 01 Jan 2016 00:00:00 +0000

Why Propensity Scores Should Not Be Used For Matching

Fri, 01 Jan 2016 00:00:00 +0000

Automating Open Science for Big Data

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

Big Data Is Not About the Data, With Applications

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

Big Data Is Not About the Data!

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

Explaining Systematic Bias and Nontransparency in US Social Security Administration Forecasts

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

Explaining Systematic Bias and Nontransparency in US Social Security Administration Forecasts

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

How Robust Standard Errors Expose Methodological Problems They Do Not Fix, and What to Do About It

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

Participant Grouping for Enhanced Interactive Experience (2nd)

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

Perusall

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

Perusall

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

Replication Data For: Explaining Systematic Bias and Nontransparency in U.S. Social Security Administration Forecasts.

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

Replication Data For: Systematic Bias and Nontransparency in U.S. Social Security Administration Forecasts.

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

Reverse-Engineering Censorship in China

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

Reverse-Engineering Censorship in China

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

Reverse-Engineering Censorship in China

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

Reverse-Engineering Censorship in China

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

Reverse-Engineering Censorship in China

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

Reverse-Engineering Censorship in China

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

RobustSE

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.

Simplifying Matching Methods for Causal Inference

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

Simplifying Matching Methods for Causal Inference

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

Simplifying Matching Methods for Causal Inference

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

System for Estimating a Distribution of Message Content Categories in Source Data (2nd)

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

Systematic Bias and Nontransparency in US Social Security Administration Forecasts

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

Talks on Matching Methods for Causal Inference

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

The Next Big [Social Science] Thing. Some Suggestions for Science Magazine

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

Urban Observatories: City Data Can Inform Decision Theory

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

Why Propensity Scores Should Not Be Used for Matching

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

Why Propensity Scores Should Not Be Used for Matching

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

Why Propensity Scores Should Not Be Used For Matching

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

Why Propensity Scores Should Not Be Used For Matching

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

Why Propensity Scores Should Not Be Used For Matching

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

Why Propensity Scores Should Not Be Used For Matching

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

Why Propensity Scores Should Not Be Used For Matching

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

An Update on Dataverse

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

Big Data Is Not About the Data!

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

Dataverse: Sharing Research Data; Building Social Science

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

Google Flu Trends Still Appears Sick: An Evaluation of the 2013‐2014 Flu Season

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

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

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

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.

Methods for Extremely Large Scale Media Experiments and Observational Studies (Poster)

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

Participant Grouping for Enhanced Interactive Experience

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

Public Policy for the Poor? A Randomized Evaluation of the Mexican Universal Health Insurance Program

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

Restructuring the Social Sciences: Reflections from Harvard's Institute for Quantitative Social Science

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

Reverse Engineering Chinese Censorship

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

Reverse Engineering Chinese Censorship

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

Reverse Engineering Chinese Censorship

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

Reverse Engineering Chinese Censorship

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

Reverse Engineering Chinese Censorship

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

Reverse Engineering Chinese Censorship

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

Reverse Engineering Chinese Censorship

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

Reverse-Engineering Censorship in China

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

Reverse-Engineering Censorship in China

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

Reverse-Engineering Censorship in China

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

Reverse-Engineering Censorship in China: Randomized Experimentation and Participant Observation

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

Simplifying Causal Inference

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

The Balance-Sample Size Frontier in Matching Methods for Causal Inference

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

The Parable of Google Flu: Traps in Big Data Analysis

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

Twitter: Big Data Opportunities—Response

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

You Lie! Patterns of Partisan Taunting in the U.S. Senate (Poster)

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

A Few IQSS Quasi-Research Projects

Tue, 01 Jan 2013 00:00:00 +0000

Big Data Is Not About the Data!

Tue, 01 Jan 2013 00:00:00 +0000

Big Data Is Not About the Data!

Tue, 01 Jan 2013 00:00:00 +0000

Big Data Is Not About the Data!

Tue, 01 Jan 2013 00:00:00 +0000

Big Data Is Not About the Data!

Tue, 01 Jan 2013 00:00:00 +0000

Big Data Is Not About the Data!

Tue, 01 Jan 2013 00:00:00 +0000

Big Data Is Not About the Data!

Tue, 01 Jan 2013 00:00:00 +0000

Big Data Is Not About the Data!

Tue, 01 Jan 2013 00:00:00 +0000

Big Data Is Not About The Data!

Tue, 01 Jan 2013 00:00:00 +0000

Big Data Is Not About The Data!

Tue, 01 Jan 2013 00:00:00 +0000

How Censorship in China Allows Government Criticism But Silences Collective Expression

Tue, 01 Jan 2013 00:00:00 +0000

How Censorship in China Allows Government Criticism But Silences Collective Expression

Tue, 01 Jan 2013 00:00:00 +0000

How Social Science Research Can Improve Teaching

Tue, 01 Jan 2013 00:00:00 +0000

Method and Apparatus for Selecting Clusterings to Classify A Predetermined Data Set

Tue, 01 Jan 2013 00:00:00 +0000

Optimizing Balance and Sample Size in Matching Methods for Causal Inference

Tue, 01 Jan 2013 00:00:00 +0000

Optimizing Balance and Sample Size in Matching Methods for Causal Inference

Tue, 01 Jan 2013 00:00:00 +0000

Public Policy for the Poor? A Randomized Evaluation of the Mexican Universal Health Insurance Program

Tue, 01 Jan 2013 00:00:00 +0000

Reverse Engineering Chinese Censorship

Tue, 01 Jan 2013 00:00:00 +0000

Reverse Engineering Chinese Censorship

Tue, 01 Jan 2013 00:00:00 +0000

Reverse Engineering Chinese Censorship

Tue, 01 Jan 2013 00:00:00 +0000

Simplifying Causal Inference

Tue, 01 Jan 2013 00:00:00 +0000

The Troubled Future of Colleges and Universities (with Comments from Five Scholar-Administrators)

Tue, 01 Jan 2013 00:00:00 +0000

What to Do About Biases in Survey Research

Tue, 01 Jan 2013 00:00:00 +0000

A Method for Computer Assisted Conceptualization

Sun, 01 Jan 2012 00:00:00 +0000

Brief of Empirical Scholars As Amici Curiae

Sun, 01 Jan 2012 00:00:00 +0000

Causal Inference Without Balance Checking: Coarsened Exact Matching

Sun, 01 Jan 2012 00:00:00 +0000

Computer-Assisted Reading'' and Other Discoveries from Quantitative Social Science

Sun, 01 Jan 2012 00:00:00 +0000

Data, Analyses, and Reports for the Arizona Independent Redistricting Commission, Filed With the U.S. Department of Justice

Sun, 01 Jan 2012 00:00:00 +0000

Analysis of Congressional Redistricting

Analysis of Legislative Redistricting

Everything above (373 MB)

Zip file

Estimating Partisan Bias of the Electoral College Under Proposed Changes in Elector Apportionment

Sun, 01 Jan 2012 00:00:00 +0000

How Censorship in China Allows Government Criticism But Silences Collective Expression

Sun, 01 Jan 2012 00:00:00 +0000

How Censorship in China Allows Government Criticism But Silences Collective Expression

Sun, 01 Jan 2012 00:00:00 +0000

Letter to the Editor on the 'Medicare Health Support Pilot Program' (by McCall and Cromwell)

Sun, 01 Jan 2012 00:00:00 +0000

Matching Methods for Causal Inference

Sun, 01 Jan 2012 00:00:00 +0000

Public Policy for the Poor? A Randomized Evaluation of the Mexican Universal Health Insurance Program

Sun, 01 Jan 2012 00:00:00 +0000

Remaking the Social Sciences

Sun, 01 Jan 2012 00:00:00 +0000

Simplifying Causal Inference

Sun, 01 Jan 2012 00:00:00 +0000

Statistical Security for Social Security

Sun, 01 Jan 2012 00:00:00 +0000

System for Estimating a Distribution of Message Content Categories in Source Data

Sun, 01 Jan 2012 00:00:00 +0000

Teaching Innovations Based on Social Science Research

Sun, 01 Jan 2012 00:00:00 +0000

Amelia II: A Program for Missing Data

Sat, 01 Jan 2011 00:00:00 +0000

An Overview of the Institute for Quantitative Social Science

Sat, 01 Jan 2011 00:00:00 +0000

Anchors: Software for Anchoring Vignettes Data

Sat, 01 Jan 2011 00:00:00 +0000

AutoCast: Automated Bayesian Forecasting With YourCast

Sat, 01 Jan 2011 00:00:00 +0000

Avoiding Randomization Failure in Program Evaluation

Sat, 01 Jan 2011 00:00:00 +0000

Comparative Effectiveness of Matching Methods for Causal Inference

Sat, 01 Jan 2011 00:00:00 +0000

Computer-Assisted Clustering and Conceptualization

Sat, 01 Jan 2011 00:00:00 +0000

Computer-Assisted Clustering and Conceptualization from Unstructured Text

Sat, 01 Jan 2011 00:00:00 +0000

Computer-Assisted Clustering and Conceptualization from Unstructured Text

Sat, 01 Jan 2011 00:00:00 +0000

Computer-Assisted Clustering and Conceptualization from Unstructured Text

Sat, 01 Jan 2011 00:00:00 +0000

Computer-Assisted Conceptualization

Sat, 01 Jan 2011 00:00:00 +0000

Computer-Assisted Conceptualization

Sat, 01 Jan 2011 00:00:00 +0000

Ensuring the Data Rich Future of the Social Sciences

Sat, 01 Jan 2011 00:00:00 +0000

Estimating Incidence Curves of Several Infections Using Symptom Surveillance Data

Sat, 01 Jan 2011 00:00:00 +0000

General Purpose Computer-Assisted Clustering and Conceptualization

Sat, 01 Jan 2011 00:00:00 +0000

Learning Catalytics (acquired by Pearson)

Sat, 01 Jan 2011 00:00:00 +0000

Matching Methods for Causal Inference

Sat, 01 Jan 2011 00:00:00 +0000

Matching Methods for Causal Inference

Sat, 01 Jan 2011 00:00:00 +0000

Matching Methods for Causal Inference

Sat, 01 Jan 2011 00:00:00 +0000

MatchIt: Nonparametric Preprocessing for Parametric Causal Inference

Sat, 01 Jan 2011 00:00:00 +0000

Multivariate Matching Methods That Are Monotonic Imbalance Bounding

Sat, 01 Jan 2011 00:00:00 +0000

Partisan Gerrymandering: A Talk and Participatory Simulation

Sat, 01 Jan 2011 00:00:00 +0000

Partisan Taunting: A New Conceptualization of How Members of Congress Act

Sat, 01 Jan 2011 00:00:00 +0000

The Future of Death in America

Sat, 01 Jan 2011 00:00:00 +0000

The Social Science Data Revolution

Sat, 01 Jan 2011 00:00:00 +0000

The Social Science Data Revolution

Sat, 01 Jan 2011 00:00:00 +0000

Topics in Measurement for the Social and Health Sciences

Sat, 01 Jan 2011 00:00:00 +0000

A General Purpose Computer-Assisted Document Clustering Methodology

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

A Hard Unsolved Problem? Post-Treatment Bias in Big Social Science Questions

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

A Method of Automated Nonparametric Content Analysis for Social Science

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

Coarsened Exact Matching

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

Comparative Effectiveness of Matching Methods for Causal Inference

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

Comparative Effectiveness of Matching Methods for Causal Inference

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

Deaths From Heart Failure: Using Coarsened Exact Matching to Correct Cause of Death Statistics

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

Designing Verbal Autopsy Studies

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

Detecting Model Dependence

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

Improving Anchoring Vignettes: Designing Surveys to Correct Interpersonal Incomparability

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

Inference in Case Control Studies

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

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

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

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).

Matching to Reduce Model Dependence

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

Public Policy for the Poor? A Randomized Evaluation of the Mexican Universal Health Insurance Program

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

Public Policy for the Poor? A Randomized Evaluation of the Mexican Universal Health Insurance Program

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

Quantitative Discovery of Qualitative Information

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

Quantitative Discovery of Qualitative Information: A General Purpose Document Clustering Methodology

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

Quantitative Discovery of Qualitative Information: A General Purpose Document Clustering Methodology

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

Quantitative Discovery of Qualitative Information: A General Purpose Document Clustering Methodology

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

ReadMe: Software for Automated Content Analysis

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

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.

What to Do About Missing Values in Time Series Cross-Section Data

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

are you making causal inferences?

Tue, 25 Aug 2009 12:00:00 +0000

Do you have a research project where you're trying to make causal inferences from observational data? Do you think matching might be a useful technique? Are you wondering how to get reviewers to stop bothering you?! Would you like some free consulting advice and data analysis help?

We're involved in some methodological research in this area and could use some experience exploring different types of data sets. If you are interested, we would be like to help you with your data analyses and inferences (for a limited number of people and a limited time). Our interactions about your data will remain between us; in particular, we promise not to scoop you, criticize you in print, or use your data for any substantive purposes at all. In fact, for most purposes we don't even need to see your dependent variable. We would be interested in reporting in our research a few aggregated statistics that test methods we are developing, but we would only do that with your permission.

If you're interested, can you send us an email?

Many thanks,

Stefano Iacus (stefano.iacus@unimi.it)
Gary King (king@harvard.edu)

Posted by Gary King at August 25, 2009 11:02 AM

A General Purpose Computer-Assisted Document Clustering Methodology

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

AMELIA II: A Program for Missing Data

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

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

Anchoring Vignettes for Interpersonal and Cross-Cultural Incomparability in Survey Research

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

CEM: Coarsened Exact Matching in Stata

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

CEM: Coarsened Exact Matching Software

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

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

CEM: Software for Coarsened Exact Matching

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

Computational Social Science

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

Discovery

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

Empirical versus Theoretical Claims about Extreme Counterfactuals: A Response

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

Extracting Systematic Social Science Meaning from Text

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

From Preserving the Past to Preserving the Future: The Data-PASS Project and the Challenges of Preserving Digital Social Science Data

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

Matched Pairs and the Future of Cluster-Randomized Experiments: A Rejoinder

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

Matching for Causal Inference Without Balance Checking

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

Matching for Causal Inference Without Balance Checking

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

OpenScholar

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

Preserving Quantitative Research-Elicited Data for Longitudinal Analysis. New Developments in Archiving Survey Data in the U.S.

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

Public Policy for the Poor? A Randomised Assessment of the Mexican Universal Health Insurance Programme

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

Public Policy for the Poor? A Randomized Evaluation of the Mexican Universal Health Insurance Program

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

Public Policy for the Poor? A Randomized Evaluation of the Mexican Universal Health Insurance Program

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

Public Policy for the Poor? A Randomized Evaluation of the Mexican Universal Health Insurance Program

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

Quantitative Discovery of Qualitative Information: A General Purpose Document Clustering Methodology

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

Quantitative Discovery of Qualitative Information: A General Purpose Document Clustering Methodology

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

Replication Data For: Public Policy for the Poor? A Randomised Assessment of the Mexican Universal Health Insurance Programme

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

The Changing Evidence Base of Social Science Research

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

The Changing Evidence Base of Social Science Research

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

The Essential Role of Pair Matching in Cluster-Randomized Experiments, With Application to the Mexican Universal Health Insurance Evaluation

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

The Future of Political Science: 100 Perspectives

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

Demographic Forecasting

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

Demographic Forecasting

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

How Not to Lie Without Statistics

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

How to Read 100 Million Blogs (& Classify Deaths Without Physicians)

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

How to Read 100 Million Blogs (& Classify Deaths Without Physicians)

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

How to Read 100 Million Blogs (& Classify Deaths Without Physicians)

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

How to Read 100 Million Blogs (& Classify Deaths Without Physicians)

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

How to Read 100 Million Blogs (& Classify Deaths Without Physicians)

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

Introduction to Matching for Causal Inference

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

Matching for Causal Inference Without Balance Checking

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

Matching for Causal Inference Without Balance Checking

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

Mexican Seguro Popular Evaluation

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

Misunderstandings Among Experimentalists and Observationalists about Causal Inference

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

Ordinary Economic Voting Behavior in the Extraordinary Election of Adolf Hitler

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

Quantitative Discovery of Qualitative Information: A General Purpose Document Clustering Methodology

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

The Dataverse Network: An Infrastructure for Data Sharing

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

The Dataverse Network: An Infrastructure for Data Sharing

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

The Dataverse Network: An Infrastructure for Data Sharing

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

The Effects of International Monetary Fund Loans on Health Outcomes

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

The Future of Death in America

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

The Future of Partisan Symmetry As a Judicial Test for Partisan Gerrymandering After LULAC V. Perry

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

Toward A Common Framework for Statistical Analysis and Development

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

VA: Verbal Autopsies

Tue, 01 Jan 2008 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.

Verbal Autopsy Methods With Multiple Causes of Death

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

A 'Politically Robust' Experimental Design for Public Policy Evaluation, With Application to the Mexican Universal Health Insurance Program

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

A 'Politically Robust' Experimental Design for Public Policy Evaluation, With Application to the Mexican Universal Health Insurance Program

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

A Proposed Standard for the Scholarly Citation of Quantitative Data

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

An Introduction to the Dataverse Network As an Infrastructure for Data Sharing

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

An Introduction to the Dataverse Network As an Infrastructure for Data Sharing

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

An Introduction to the Dataverse Network As an Infrastructure for Data Sharing

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

An Introduction to the Dataverse Network As an Infrastructure for Data Sharing

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

Anchoring Vignettes for Interpersonally Incomparable Survey Responses

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

Anchors: Software for Anchoring Vignettes Data

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

Anchors: Software for Anchoring Vignettes Data

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

Comparing Incomparable Survey Responses: New Tools for Anchoring Vignettes

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

Crimson Hexagon (merged With Brandwatch, Acquired by Cision)

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

Dataverse: Open Source Research Data Repository Software

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

Dataverse: Open Source Research Data Repository Software

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

Detecting Model Dependence in Statistical Inference: A Response

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

Extracting Systematic Social Science Meaning from Text (& Cause-Specific Mortality Rates from Symptom Data)

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

How to Read 100 Million Blogs (& Classify Deaths Without Physicians)

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

How to Read 100 Million Blogs (& Classify Deaths Without Physicians)

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

How to Read 100 Million Blogs (& Classify Deaths Without Physicians)

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

How to Read 100 Million Blogs (& Classify Deaths Without Physicians)

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

How to Read 100 Million Blogs (& Classify Deaths Without Physicians)

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

How to Read 100 Million Blogs (& Classify Deaths Without Physicians)

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

Matching As Nonparametric Preprocessing for Reducing Model Dependence in Parametric Causal Inference

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

MatchIt: Nonparametric Preprocessing for Parametric Causal Inference

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

MatchIt: Nonparametric Preprocessing for Parametric Causal Inference

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

Understanding the Lee-Carter Mortality Forecasting Method

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

What to Do about Biases in Survey Research

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

When Can History Be Our Guide? The Pitfalls of Counterfactual Inference

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

A 'Politically Robust' Experimental Design for Public Policy Evaluation, With Application to the Mexican Universal Health Insurance Program

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

Death by Survey: Estimating Adult Mortality Without Selection Bias from Sibling Survival Data

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

Demographic Forecasting

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

Demographic Forecasting

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

Detecting and Reducing Model Dependence in Causal Inference

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

Detecting and Reducing Model Dependence in Causal Inference

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

Ecological Inference

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

Evaluation of Seguro Popular: Baseline Analysis

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

Finding, Citing, Analyzing, Disseminating, and Preserving Quantitative Data

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

Publication, Publication

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

Statistical Analysis With Missing Data

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

The Dangers of Extreme Counterfactuals

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

The Effect of War on the Supreme Court

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

Zelig: Everyone's Statistical Software

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

Zelig: Everyone's Statistical Software

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

The Value of Control Groups in Causal Inference (and Breakfast Cereal)

Mon, 31 Oct 2005 12:00:00 +0000

A few years ago, I taught the following lesson in my daughter's kindergarden class and my graduate methods class in the same week. It worked pretty well in both. Anyone who has a kid in kindergarten, some good graduate students, or both, might want to try this. It was especially fun for the instructor.

To start, I hold up some nails and ask "does everyone likes to eat nails?" The kindergarten kids scream, "Nooooooo." The graduate students say "No," trying to look cool. I say I'm going to convince them otherwise.

I hand out a little magnet to everyone. I ask the class to figure out what it sticks to and what it doesn't stick to. After a few minutes running around the classroom, the kindergardners figure out that magnets stick to stuff with iron in it, and anything without iron in it doesn't stick. The graduate students sit there looking cool.

From behind the table, I pull out a box of Total Cereal (in my experience, teaching is just like doing magic tricks, except that you get paid more as a magician). I show them the list of ingredients; "iron, 100 percent" is on the list. I ask by a show of hands whether this is the same iron as in the nails. 3 of 23 kindergarten kids say "yes"; 5 of 44 Harvard graduate students say "yes" (almost the same percent in both classes!).

I show the students that the box is sealed (and I have nothing up my sleeves), Then, I open the box, spill some cereal on a table, and smash it up into tiny pieces with a rolling pin. I take the pile of squashed cereal around the room and let the kids put their magnet next to it and see whether the cereal sticks to the magnet. To everyone's amazement, it sticks!

Then I ask, "are we now convinced that the iron in the nails is the same iron as in the cereal?" All the kids in kindergarten and all the graduate students say "yes."

I respond by saying "but how do you know the cereal stuck to the magnet because it had iron in it? Maybe it was just sticky, like gum or tape." Now that I finally have their attention (not a minor matter with kindergartners), I get to explain to them what a control group is (the point of the lesson). And from behind the table, I pull out a box of Rice Krispies (which are made of nothing). We examine the side of the box to verify the lack of (much) iron, and then I smash up the Rice Krispies, and let them see if their magnet sticks. It doesn't stick!

Everyone gets to take home a cool fact (they love to eat the stuff in nails), I get to convey the point of the lesson in a way they won't forget (the essential role of control groups in causal inference), and everyone gets a free magnet.

(This post was originally published on 10/31/2005. Since then, Kellogg's started to put iron in Rice Krispies, so to do this experiment you now need to find some other cereal. I find that cereal marked "organic" often doesn't have added iron.)

A Social Science of Architecture

Tue, 18 Oct 2005 12:00:00 +0000

After eight years of learning something about architecture (from Harry Cobb and his team) and extensive programmatic planning, the Institute for Quantitative Social Science this semester moves into the new Center for Government and International Studies buildings. Our official address is the Third Floor of 1737 Cambridge Street (the design is vaguely reminiscent of the bridge of the Starship Enterprise), although we also occupy some of the other floors and some of the building across the street. It is not really finished yet, but it is a terrific facility, with floor to ceiling windows in most offices, a wonderful seminar room for our Applied Statistics Workshop, and many other useful features. Perhaps even more remarkably, everyone seems to love it (Congratulations Harry!).

One issue I learned during this long process was how the field of architecture has the best science, engineering, and art, but very little modern social scientific analysis. Yet, social science, quantitative social science in particular, could greatly help architecture achieve its goals, I think. Ultimately the goal of this particular $100M-plus building, and of most buildings built by universities, is not only to create beautiful surroundings but also to increase the amount of knowledge created, disseminated, and preserved (my summary of the purpose of modern research universities). So do not limit yourself to asking how a building makes you feel, what architectural critics might think, how it fits in with the style of other buildings on campus, or whether your office is to your liking. Ask instead, or in addition, whether the building increases the units of knowledge created, disseminated, and preserved more than some other building or some other potential use for the money. This strikes me as the central question to be answered by those who decide what buildings to build, and yet the systematic scientific basis for this decision is almost nonexistent.

As such, some systematic data collection could have a considerable impact on this field. Do corridors or suites make the faculty and students produce and learn more? Does vertical circulation work as well as horizontal? Should we put faculty in close proximity to others working on the same projects or should we maximize interdisciplinary adjacencies? Which types of floor plans increase interaction? Which types of interaction produce the most knowledge created, generated, and preserved? Do we want to build buildings that encourage doors to be kept open, so as to make the faculty seem approachable or should we try to keep doors closed so that they can get work done? In this field as in most others, a great deal can be learned by directly measuring the relevant outcome variable; in architecture, quite remarkably, this has only rarely been attempted.

Of course it is done all the time via qualitative judgments, but in almost every field of science where a sufficient fraction of information can be quantified, statistical analysis beats human judgment. There is no reason to think that the same kind of statistical science wouldn't also create enormous advances here too.

I have heard of a couple of isolated academic works on this subject, but we're talking about some of the most important and expensive decisions universities make (and among the biggest decisions businesses, and many other institutions make too). There should be an entire subfield devoted to the subject. All it would take is some data collection and analysis. Outcome measures could include, for example faculty citation rates, publications, awards, grants, and departmental rankings, along with student recruitment, retention, graduation, and placement rates. The key treatment variables would include various information on the types of buildings and architectural design. Random assignment seems infeasible, but relatively exogenous features might include departmental moves or city and town building restrictions. Universities that allow faculty the choice of buildings could also provide useful revealed preference measures. I would think that a few enterprising scholars on this path could have an enormous impact both in creating a new academic subfield and in improving a vitally important set of university (and societal) decisions.

In the interm, we'll enjoy the new buildings and hope they have a positive impact.

Brief of Amici Curiae Professors Gary King, Bernard Grofman, Andrew Gelman, and Jonathan Katz in Support of Neither Party

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

Death by Survey: Estimating Adult Mortality Without Selection Bias from Sibling Survival Data

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

Demographic Forecasting

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

Demographic Forecasting

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

Demographic Forecasting

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

Demographic Forecasting: : Incorporating Qualitative Insight into Quantitative Modeling

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

Evaluation of Seguro Popular: Design Framework

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

Finding, Citing, Analyzing, Disseminating, and Preserving Numeric Data

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

Model Dependence in Counterfactual Inference

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

The Supreme Court During Crisis: How War Affects only Non-War Cases

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

WhatIf: Software for Evaluating Counterfactuals

Sat, 01 Jan 2005 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

Did Illegal Overseas Absentee Ballots Decide the 2000 U.S. Presidential Election?

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

Ecological Inference: New Methodological Strategies

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

EI: A Program for Ecological Inference

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

Empirically Evaluating the Electoral College

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

Enhancing the Validity and Cross-Cultural Comparability of Measurement in Survey Research

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

Finding New Information for Ecological Inference Models: A Comment on Jon Wakefield, 'Ecological Inference in 2X2 Tables'

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

Excerpt:

Congratulations goes to Jon Wakefield for an unusually complete and completely insightful contribution to this fast-growing literature. Wakefield productively follows what is now standard practice by including both deterministic and statistical information in each new model and then seeking out additional sources of information.

Gill/Murray/Cholesky/Factorization

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

Gill/Murray/Cholesky/Factorization

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

Inference in Case-Control Studies

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

Information in Ecological Inference: An Introduction

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

Schnabel/Eskow/Cholesky/Factorization

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

Schnabel/Eskow/Cholesky/Factorization

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

Theory and Evidence in International Conflict: A Response to de Marchi, Gelpi, and Grynaviski

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

What to Do When Your Hessian Is Not Invertible: Alternatives to Model Respecification in Nonlinear Estimation

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

YourCast

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

10 Million International Dyadic Events

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

A Consensus on Second Stage Analyses in Ecological Inference Models

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

An Automated Information Extraction Tool For International Conflict Data With Performance As Good As Human Coders: A Rare Events Evaluation Design

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

Analyzing Second Stage Ecological Regressions

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

Building An Infrastructure for Empirical Research in the Law

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

CLARIFY: Software for Interpreting and Presenting Statistical Results

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

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

Determinants of Inequality in Child Survival: Results from 39 Countries

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

EI: A Program for Ecological Inference

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

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

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

Numerical Issues Involved in Inverting Hessian Matrices

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

ReLogit: Rare Events Logistic Regression

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

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.

Some Statistical Methods for Evaluating Information Extraction Systems

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

The Future of Replication

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

A Fast, Easy, and Efficient Estimator for Multiparty Electoral Data

Tue, 01 Jan 2002 00:00:00 +0000

Archive of the Controversy Involving Wendy K. Tam Cho, Brian J. Gaines, and the American Political Science Review

Tue, 01 Jan 2002 00:00:00 +0000

Armed Conflict As a Public Health Problem

Tue, 01 Jan 2002 00:00:00 +0000

COUNT: A Program for Estimating Event Count and Duration Regressions

Tue, 01 Jan 2002 00:00:00 +0000

Empirical Research and The Goals of Legal Scholarship: A Response

Tue, 01 Jan 2002 00:00:00 +0000

Estimating Risk and Rate Levels, Ratios, and Differences in Case-Control Studies

Tue, 01 Jan 2002 00:00:00 +0000

Isolating Spatial Autocorrelation, Aggregation Bias, and Distributional Violations in Ecological Inference

Tue, 01 Jan 2002 00:00:00 +0000

Measuring Total Health Inequality: Adding Individual Variation to Group-Level Differences

Tue, 01 Jan 2002 00:00:00 +0000

Rethinking Human Security

Tue, 01 Jan 2002 00:00:00 +0000

The Rules of Inference

Tue, 01 Jan 2002 00:00:00 +0000

A Digital Library for the Dissemination and Replication of Quantitative Social Science Research

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

Aggregation Among Binary, Count, and Duration Models: Estimating the Same Quantities from Different Levels of Data

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

An Introduction to the Virtual Data Center Project and Software

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

An Overview of the Virtual Data Center Project and Software

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

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.

Analyzing Incomplete Political Science Data: An Alternative Algorithm for Multiple Imputation

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

Bayesian and Frequentist Inference for Ecological Inference: The RxC Case

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

Explaining Rare Events in International Relations

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

Improving Forecasts of State Failure

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

Logistic Regression in Rare Events Data

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

Proper Nouns and Methodological Propriety: Pooling Dyads in International Relations Data

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

Geography, Statistics, and Ecological Inference

Sat, 01 Jan 2000 00:00:00 +0000

Improving Quantitative Studies of International Conflict: A Conjecture

Sat, 01 Jan 2000 00:00:00 +0000

Making the Most of Statistical Analyses: Improving Interpretation and Presentation

Sat, 01 Jan 2000 00:00:00 +0000

A Statistical Model for Multiparty Electoral Data

Fri, 01 Jan 1999 00:00:00 +0000

Binomial-Beta Hierarchical Models for Ecological Inference

Fri, 01 Jan 1999 00:00:00 +0000

Many Publications, But Still No Evidence

Fri, 01 Jan 1999 00:00:00 +0000

No Evidence on Directional Vs. Proximity Voting

Fri, 01 Jan 1999 00:00:00 +0000

Not Asked and Not Answered: Multiple Imputation for Multiple Surveys

Fri, 01 Jan 1999 00:00:00 +0000

The Future of Ecological Inference Research: A Reply to Freedman et Al.

Fri, 01 Jan 1999 00:00:00 +0000

AMELIA: A Program for Missing Data

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

Estimating the Probability of Events That Have Never Occurred: When Is Your Vote Decisive?

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

MAXLIK

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

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.

The Record of American Democracy, 1984-1990

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

Unifying Political Methodology: The Likelihood Theory of Statistical Inference

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

A Solution to the Ecological Inference Problem: Reconstructing Individual Behavior from Aggregate Data

Wed, 01 Jan 1997 00:00:00 +0000

A Preview of EI and EzI: Programs for Ecological Inference

Mon, 01 Jan 1996 00:00:00 +0000

Advantages of Conflictual Redistricting

Mon, 01 Jan 1996 00:00:00 +0000

Racial Fairness in Legislative Redistricting

Mon, 01 Jan 1996 00:00:00 +0000

The Generalization in the Generalized Event Count Model, With Comments on Achen, Amato, and Londregan

Mon, 01 Jan 1996 00:00:00 +0000

Why Context Should Not Count

Mon, 01 Jan 1996 00:00:00 +0000

A Correction for an Underdispersed Event Count Probability Distribution

Sun, 01 Jan 1995 00:00:00 +0000

A Revised Proposal, Proposal

Sun, 01 Jan 1995 00:00:00 +0000

Pre-Election Survey Methodology: Details From Nine Polling Organizations, 1988 and 1992

Sun, 01 Jan 1995 00:00:00 +0000

Replication, Replication

Sun, 01 Jan 1995 00:00:00 +0000

The Importance of Research Design in Political Science

Sun, 01 Jan 1995 00:00:00 +0000

A Unified Method of Evaluating Electoral Systems and Redistricting Plans

Sat, 01 Jan 1994 00:00:00 +0000

Designing Social Inquiry: Scientific Inference in Qualitative Research

Sat, 01 Jan 1994 00:00:00 +0000

Elections to the United States House of Representatives, 1898-1992

Sat, 01 Jan 1994 00:00:00 +0000

Enhancing Democracy Through Legislative Redistricting

Sat, 01 Jan 1994 00:00:00 +0000

Party Competition and Media Messages in U.S. Presidential Election Campaigns

Sat, 01 Jan 1994 00:00:00 +0000

Transfers of Governmental Power: The Meaning of Time Dependence

Sat, 01 Jan 1994 00:00:00 +0000

Good Research and Bad Research: Extending Zimile's Criticism

Fri, 01 Jan 1993 00:00:00 +0000

On Party Platforms, Mandates, and Government Spending

Fri, 01 Jan 1993 00:00:00 +0000

The Methodology of Presidential Research

Fri, 01 Jan 1993 00:00:00 +0000

The Science of Political Science Graduate Admissions

Fri, 01 Jan 1993 00:00:00 +0000

Why Are American Presidential Election Campaign Polls so Variable When Votes Are so Predictable?

Fri, 01 Jan 1993 00:00:00 +0000

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

Wed, 01 Jan 1992 00:00:00 +0000

'Truth' Is Stranger Than Prediction, More Questionable Than Causal Inference

Tue, 01 Jan 1991 00:00:00 +0000

Calculating Standard Errors of Predicted Values Based on Nonlinear Functional Forms

Tue, 01 Jan 1991 00:00:00 +0000

Constituency Service and Incumbency Advantage

Tue, 01 Jan 1991 00:00:00 +0000

On Political Methodology

Tue, 01 Jan 1991 00:00:00 +0000

Stochastic Variation: A Comment on Lewis-Beck and Skalaban's 'The R-Square'

Tue, 01 Jan 1991 00:00:00 +0000

Systemic Consequences of Incumbency Advantage in the U.S. House

Tue, 01 Jan 1991 00:00:00 +0000

A Unified Model of Cabinet Dissolution in Parliamentary Democracies

Mon, 01 Jan 1990 00:00:00 +0000

Electoral Responsiveness and Partisan Bias in Multiparty Democracies

Mon, 01 Jan 1990 00:00:00 +0000

Estimating Incumbency Advantage Without Bias

Mon, 01 Jan 1990 00:00:00 +0000

Estimating the Electoral Consequences of Legislative Redistricting

Mon, 01 Jan 1990 00:00:00 +0000

Measuring the Consequences of Delegate Selection Rules in Presidential Nominations

Mon, 01 Jan 1990 00:00:00 +0000

A Seemingly Unrelated Poisson Regression Model

Sun, 01 Jan 1989 00:00:00 +0000

Electoral Responsiveness in U.S. Congressional Elections, 1946-1986

Sun, 01 Jan 1989 00:00:00 +0000

Event Count Models for International Relations: Generalizations and Applications

Sun, 01 Jan 1989 00:00:00 +0000

Representation Through Legislative Redistricting: A Stochastic Model

Sun, 01 Jan 1989 00:00:00 +0000

The Presidency in American Politics

Sun, 01 Jan 1989 00:00:00 +0000

Variance Specification in Event Count Models: From Restrictive Assumptions to a Generalized Estimator

Sun, 01 Jan 1989 00:00:00 +0000

Statistical Models for Political Science Event Counts: Bias in Conventional Procedures and Evidence for The Exponential Poisson Regression Model

Fri, 01 Jan 1988 00:00:00 +0000

The Elusive Executive: Discovering Statistical Patterns in the Presidency

Fri, 01 Jan 1988 00:00:00 +0000

Democratic Representation and Partisan Bias in Congressional Elections

Thu, 01 Jan 1987 00:00:00 +0000

Presidential Appointments to the Supreme Court: Adding Systematic Explanation to Probabilistic Description

Thu, 01 Jan 1987 00:00:00 +0000

Seats, Votes, and Gerrymandering: Measuring Bias and Representation in Legislative Redistricting

Thu, 01 Jan 1987 00:00:00 +0000

How Not to Lie With Statistics: Avoiding Common Mistakes in Quantitative Political Science

Wed, 01 Jan 1986 00:00:00 +0000

Political Parties and Foreign Policy: A Structuralist Approach

Wed, 01 Jan 1986 00:00:00 +0000

The Development of Political Activists: A Model of Early Learning

Wed, 01 Jan 1986 00:00:00 +0000

The Significance of Roll Calls in Voting Bodies: A Model and Statistical Estimation

Wed, 01 Jan 1986 00:00:00 +0000

Book Review of `Forecasting Presidential Elections'

Tue, 01 Jan 1985 00:00:00 +0000

The Stability of Party Identification Among U.S. Representatives: Political Loyalty, 1789-1984

Sun, 01 Jan 1984 00:00:00 +0000

PC$: Checkbook Manager

Fri, 01 Jan 1982 00:00:00 +0000

The City's Losing Clout

Mon, 01 Jan 1979 00:00:00 +0000

Aaron Kaufman

Mon, 01 Jan 0001 00:00:00 +0000

Abhradeep Guha Thakurta

Mon, 01 Jan 0001 00:00:00 +0000

Accessibility Statement

Mon, 01 Jan 0001 00:00:00 +0000

This site is part of Harvard University and works to support the Harvard University Digital Accessibility Policy, which adopts the W3C Web Content Accessibility Guidelines (WCAG) 2.1, Level AA as its standard.

Our commitment

We aim to make the content on this site usable by everyone, including people who use assistive technologies such as screen readers, screen magnification, alternative keyboards, voice input, or who navigate the site without a mouse. Toward that goal we work to ensure that the site:

Provides text alternatives for non-text content (images, icons).
Uses semantic HTML landmarks (<header>, <nav>, <main>, <footer>) and a “Skip to main content” link so keyboard users can bypass repeated navigation.
Maintains color and text contrast that meets WCAG 2.1 AA.
Provides visible keyboard focus indicators on links, buttons, and form controls, and supports keyboard navigation of menus and dialogs.
Labels every form control programmatically (not only via placeholder text).
Uses headings in a logical order to communicate document structure.
Uses descriptive link text and clearly identifies links that open in a new browser tab.

Known limitations

Some legacy publications archived on this site (older PDFs, scanned articles, and historical slides) may not yet be fully accessible. We work to remediate them as resources allow, and prioritize widely used content first.

If you encounter content that is not accessible, an accessible alternative can typically be made available on request.

Reporting an issue or requesting an alternative format

If you have difficulty accessing any material on this site, or if you would like content provided in an alternative accessible format, please contact us:

Use the accessibility report form maintained by Harvard University Information Technology, or
Reach out via the Contact page and we will respond as quickly as possible.

Standards & resources

Last reviewed: April 2026

Ajay Tandon

Mon, 01 Jan 0001 00:00:00 +0000

Akshay Swaminathan

Mon, 01 Jan 0001 00:00:00 +0000

Alan Lopez

Mon, 01 Jan 0001 00:00:00 +0000

Albert Pereta

Mon, 01 Jan 0001 00:00:00 +0000

Albert-László Barabási

Mon, 01 Jan 0001 00:00:00 +0000

Aleksandra Conevska

Mon, 01 Jan 0001 00:00:00 +0000

Alessandro Vespignani

Mon, 01 Jan 0001 00:00:00 +0000

Alex Pentland

Mon, 01 Jan 0001 00:00:00 +0000

Alexander Schuessler

Mon, 01 Jan 0001 00:00:00 +0000

Alexander Wagner

Mon, 01 Jan 0001 00:00:00 +0000

Alexis Diamond

Mon, 01 Jan 0001 00:00:00 +0000

Allen Schmaltz

Mon, 01 Jan 0001 00:00:00 +0000

Alok Kumar

Mon, 01 Jan 0001 00:00:00 +0000

Alondra Nelson

Mon, 01 Jan 0001 00:00:00 +0000

Amac Herdagdelen

Mon, 01 Jan 0001 00:00:00 +0000

Amher Tarar

Mon, 01 Jan 0001 00:00:00 +0000

Amine Kamel

Mon, 01 Jan 0001 00:00:00 +0000

Anders Corr

Mon, 01 Jan 0001 00:00:00 +0000

Andrew B. Hall

Mon, 01 Jan 0001 00:00:00 +0000

Andrew C Thomas

Mon, 01 Jan 0001 00:00:00 +0000

Andrew C. Eggers

Mon, 01 Jan 0001 00:00:00 +0000

Andrew Gelman

Mon, 01 Jan 0001 00:00:00 +0000

Andrew Reeves

Mon, 01 Jan 0001 00:00:00 +0000

Andy Shi

Mon, 01 Jan 0001 00:00:00 +0000

Ann Joseph

Mon, 01 Jan 0001 00:00:00 +0000

Anton Strezhnev

Mon, 01 Jan 0001 00:00:00 +0000

Antonio Câmara

Mon, 01 Jan 0001 00:00:00 +0000

Ardys Kozbial

Mon, 01 Jan 0001 00:00:00 +0000

Are you applying to a Ph.D., MA, or Post-Doc program at Harvard?

Mon, 01 Jan 0001 00:00:00 +0000

The Harvard Government Department often receives 800+ applications a year, and similarly for other Harvard programs I'm involved in. As such, meeting with prospective students is, I regret, infeasible. [If we all did this, we wouldn't have time to teach the students already here, our students would leave, few would apply to attend, and we'd then have plenty of time to meet with you. But of course then you wouldn't want to come!] A better plan is to contact us after you're admitted.

In some Ph.D. programs (in the natural and physical sciences and in the social sciences in other countries), students are admitted to study with a specific individual faculty member who often makes the choice about admission; this is not the case in our program, where decisions are made by an admissions committee on behalf of the entire department. So you don't need to contact me to ask whether you can study with me; just apply to the department and if you get in you can.

You might be interested in this article I wrote about the admissions process in our department: "The Science of Political Science Graduate Admissions". The odds of being admitted are not high for anyone, but they do have to admit someone, and it might as well be you! I'm also happy to report, as a proud graduate of the University of Wisconsin, that there are many other great programs out there. Best of luck with the application process.

Ariel R. White

Mon, 01 Jan 0001 00:00:00 +0000

Aristides A. N. Patrinos

Mon, 01 Jan 0001 00:00:00 +0000

Aslaug Asgeirsdottir

Mon, 01 Jan 0001 00:00:00 +0000

Atul Gawande

Mon, 01 Jan 0001 00:00:00 +0000

Avleen S. Bijral

Mon, 01 Jan 0001 00:00:00 +0000

Aykut Firat

Mon, 01 Jan 0001 00:00:00 +0000

Beau Coker

Mon, 01 Jan 0001 00:00:00 +0000

Ben Bishin

Mon, 01 Jan 0001 00:00:00 +0000

Ben Silbermann

Mon, 01 Jan 0001 00:00:00 +0000

Ben W. Hunt

Mon, 01 Jan 0001 00:00:00 +0000

Benjamin Schneer

Mon, 01 Jan 0001 00:00:00 +0000

Bernard Grofman

Mon, 01 Jan 0001 00:00:00 +0000

Bernard Tamas

Mon, 01 Jan 0001 00:00:00 +0000

Bhalachandra S. Kodkany

Mon, 01 Jan 0001 00:00:00 +0000

Bharath Kumar

Mon, 01 Jan 0001 00:00:00 +0000

Bio & CV

Mon, 01 Jan 0001 00:00:00 +0000

Gary King is the Albert J. Weatherhead III University Professor at Harvard University – one of 25 with Harvard’s most distinguished faculty title – and Director of the Institute for Quantitative Social Science. King develops and applies empirical methods in many areas of social science, focusing on innovations that span the range from statistical theory to practical application.

King is an elected Fellow in 8 honorary societies (National Academy of Sciences, American Statistical Association, American Association for the Advancement of Science, American Academy of Arts and Sciences, Society for Political Methodology, National Academy of Social Insurance, American Academy of Political and Social Science, and the Guggenheim Foundation) and has won more than 55 prizes and awards for his work. King was elected President of the Society for Political Methodology and Vice President of the American Political Science Association. He has been a member of the Senior Editorial Board at Science, Visiting Fellow at Oxford, and Senior Science Adviser to the World Health Organization. He has written more than 190 journal articles, 30 open source software packages, and 8 books.

King proposed the now widely accepted standard for fairness in legislative redistricting known as “partisan symmetry,” and the methods used by courts and parties to detect when partisan gerrymandering violates it. His “ecological inference” methods for inferring individual behavior from aggregate data are used in most jurisdictions applying the Voting Rights Act to detect racial gerrymandering. His book with Keohane and Verba, Designing Social Inquiry, helped launch the modern subfield of qualitative methods in political science; his book Unifying Political Methodology had a similar role for quantitative political methodology. His “Replication, Replication” article helped initiate the data sharing movement in political science, and his ongoing international “Dataverse” project supports the movement across fields. His “anchoring vignettes” approach to cross-cultural survey comparability has been used in more than 100 countries by researchers, governments, and others. King has pioneered “politically robust” research designs that make possible unusually large randomized experiments in politically difficult circumstances – including the largest ever randomized health policy experiment, to evaluate the Mexican universal health insurance program, and the only large scale randomized news media experiment in the US. He has reverse engineered Chinese censorship and fabrication of social media posts, improved Social Security Trust Fund forecasts, and developed empirical methods and software widely used in academia, government, and private industry for automated text analysis, rare events, missing data, measurement error, causal inference, interpreting statistical results, and for forecasting elections, mortality rates, and international conflict.

King’s work is widely read across scholarly fields and beyond academia. He was listed as the most cited political scientist of his cohort; among the group of “political scientists who have made the most important theoretical contributions” to the discipline “from its beginnings in the late-19th century to the present”; and on lists of the most highly cited researchers across the social sciences. King’s many former students and postdocs now hold positions at leading universities and companies around the world. He has collaborated with hundreds of scholars, including many of his students, on research for publication. He has served on more than 30 editorial, nonprofit, and corporate boards; as founding editor of The Political Methodologist, and on the governing councils of the American Political Science Association, Inter-university Consortium for Political and Social Research, Society for Political Methodology, Midwest Political Science Association, Center for the Advanced Study in the Behavioral Sciences, and Institute for Data, Science, and Society.

King is also co-founder and an inventor of the original technology for Crimson Hexagon (merged with Brandwatch, acquired by Cision), Learning Catalytics (acquired by Pearson), Thresher (acquired by Two Six Technologies, a Carlyle Company), OpenScholar (acquired by Monomyth Group), Perusall, and QuickCode. He has received 17 patents for these technologies.

King is a proud graduate of SUNY New Paltz (B.A., 1980) and the University of Wisconsin-Madison (M.A., Ph.D., 1984). He taught at NYU for three years before coming to Harvard in 1987.

Download CV (PDF)

Bradley Palmquist

Mon, 01 Jan 0001 00:00:00 +0000

Brady Baybeck

Mon, 01 Jan 0001 00:00:00 +0000

Brandon M. Stewart

Mon, 01 Jan 0001 00:00:00 +0000

Brenda Sequeira D'Mello

Mon, 01 Jan 0001 00:00:00 +0000

Brent A. Coull

Mon, 01 Jan 0001 00:00:00 +0000

Brian Lukoff

Mon, 01 Jan 0001 00:00:00 +0000

Carlos Velasco Rivera

Mon, 01 Jan 0001 00:00:00 +0000

Casey Petroff

Mon, 01 Jan 0001 00:00:00 +0000

Casey S. Greene

Mon, 01 Jan 0001 00:00:00 +0000

Chengyu Fu

Mon, 01 Jan 0001 00:00:00 +0000

Chiara Superti

Mon, 01 Jan 0001 00:00:00 +0000

Chris Danford

Mon, 01 Jan 0001 00:00:00 +0000

Christina Lee

Mon, 01 Jan 0001 00:00:00 +0000

Christine Harrington

Mon, 01 Jan 0001 00:00:00 +0000

Christopher Adolph

Mon, 01 Jan 0001 00:00:00 +0000

Christopher Bingham

Mon, 01 Jan 0001 00:00:00 +0000

Christopher Lucas

Mon, 01 Jan 0001 00:00:00 +0000

Christopher Murray

Mon, 01 Jan 0001 00:00:00 +0000

Christopher T. Kenny

Mon, 01 Jan 0001 00:00:00 +0000

Chuanhai Liu

Mon, 01 Jan 0001 00:00:00 +0000

Claudia Pedroza

Mon, 01 Jan 0001 00:00:00 +0000

Claudia Wagner

Mon, 01 Jan 0001 00:00:00 +0000

Claudine Gay

Mon, 01 Jan 0001 00:00:00 +0000

Clayton Nall

Mon, 01 Jan 0001 00:00:00 +0000

Colin Fredericks

Mon, 01 Jan 0001 00:00:00 +0000

Connor D. Huff

Mon, 01 Jan 0001 00:00:00 +0000

Connor T. Jerzak

Mon, 01 Jan 0001 00:00:00 +0000

Contact

Mon, 01 Jan 0001 00:00:00 +0000

Gary King

Institute for Quantitative Social Science 1737 Cambridge Street Harvard University Cambridge, MA 02138

Web: GaryKing.org

Twitter: @KingGary

Assistant: Maria Martins, (617) 495-9271
Email: king-assist@iq.harvard.edu

Signal, Phone: 617, five hundred, 75 seventy

Email: King@Harvard.edu

ORCID: https://orcid.org/0000-0002-5327-7631

Cory McCartan

Mon, 01 Jan 0001 00:00:00 +0000

Curtis Signorino

Mon, 01 Jan 0001 00:00:00 +0000

Cynthia Dwork

Mon, 01 Jan 0001 00:00:00 +0000

Cynthia Rudin

Mon, 01 Jan 0001 00:00:00 +0000

D. Stephen Voss

Mon, 01 Jan 0001 00:00:00 +0000

Dan Ho

Mon, 01 Jan 0001 00:00:00 +0000

Dan Ward

Mon, 01 Jan 0001 00:00:00 +0000

Dana Higgins

Mon, 01 Jan 0001 00:00:00 +0000

Daniel Gilbert

Mon, 01 Jan 0001 00:00:00 +0000

Daniel Hopkins

Mon, 01 Jan 0001 00:00:00 +0000

Daniel J. Walsh

Mon, 01 Jan 0001 00:00:00 +0000

Daniel K. Baissa

Mon, 01 Jan 0001 00:00:00 +0000

Daniel L Kiskis

Mon, 01 Jan 0001 00:00:00 +0000

Danielle E. Tuller

Mon, 01 Jan 0001 00:00:00 +0000

Danny Ebanks

Mon, 01 Jan 0001 00:00:00 +0000

Dataverse

Mon, 01 Jan 0001 00:00:00 +0000

Below is my Dataverse collection, which is comprised of data sets and replication data sets associated with my published articles and books. For information about the Dataverse project (which I created and run), see this article and the Dataverse.org project website.

David (Wendong) Zhang

Mon, 01 Jan 0001 00:00:00 +0000

Undergraduate student in the Department of Economics at Harvard University. He is currently exploring interdisciplinary interests in statistics and computer science.

David Kane

Mon, 01 Jan 0001 00:00:00 +0000

David Lazer

Mon, 01 Jan 0001 00:00:00 +0000

David Leal

Mon, 01 Jan 0001 00:00:00 +0000

David Lublin

Mon, 01 Jan 0001 00:00:00 +0000

David R. Cheng

Mon, 01 Jan 0001 00:00:00 +0000

Dawn Brancati

Mon, 01 Jan 0001 00:00:00 +0000

Deb Roy

Mon, 01 Jan 0001 00:00:00 +0000

Debra Javeline

Mon, 01 Jan 0001 00:00:00 +0000

Deen Freelon

Mon, 01 Jan 0001 00:00:00 +0000

Delia Bailey

Mon, 01 Jan 0001 00:00:00 +0000

Dennis M. Feehan

Mon, 01 Jan 0001 00:00:00 +0000

Designing Political Inquiry, Government 1003 (for undergrads)

Mon, 01 Jan 0001 00:00:00 +0000

Not offered this year.

Class Materials:

Syllabus for Designing Political Research, G1003

This class is for students who wish to learn how to conduct, and evaluate, modern social science research. For students planning to write senior theses; considering graduate school; who would like to understand the concept of ``evidence'' for law school; or thinking about taking a job with a consulting firm, research is almost the only skill you need to learn. The goal will be to assess the state of a scholarly literature, identify the interesting questions, formulate strategies for answering them, acquire the methodological tools with which to conduct the research, and understand how to write up the results so they can be published.

Although most undergraduate, and even most graduate courses, address these issues indirectly, we provide an explicit analysis of each. We do this in the context of a variety of strategies of empirical political inquiry. Our examples cover American politics, international relations, comparative politics, and other subfields of political science and social science that rely on empirical evidence. We do not address certain research in political theory for which empirical evidence is not central, but our methodological emphases will be as varied as our substantive examples. We take empirical evidence to be historical, quantitative, or anthropological. Specific methodologies include survey research, experiments, non-experiments, intensive interviews, statistical analyses, case studies, and participant observation.

Devon Brewer

Mon, 01 Jan 0001 00:00:00 +0000

Do You Have a Paper to Submit to PNAS?

Mon, 01 Jan 0001 00:00:00 +0000

The Proceedings of the National Academy of Sciences has unusual submission procedures. Most submissions go through Direct Review, which is similar to the traditional review process at most scholarly journals, except that “Authors must recommend three appropriate Editorial Board members, three NAS members who are expert in the paper’s scientific area, and five qualified reviewers.” Alternatively, authors may ask any elected member of the National Academy of Sciences to serve as a Prearranged Editor (PE), who will then lead the review process. Although the procedures say that “PEs should be used only when an article falls into an area without broad representation in the Academy, or for research that may be considered counter to a prevailing view or too far ahead of its time to receive a fair hearing, and in which the member is expert,” defining these boundaries is difficult.

Although I respect the role of journal editor and deeply appreciate and honor those who take on this crucial job in our profession, I prefer to make my contributions in other ways. So I normally decline invitations to serve as PE (or member-editor, which is another category) for PNAS (and as editor, associate editor, etc., for other journals). If you think I’d be appropriate, I’d be more than happy to serve as reviewer, as I do write numerous reviews. In addition, I have an arrangement with members of the PNAS editorial board in subfields I touch on where I take on some aspects of the editorial role by recommending scholars to serve as reviewers and editors, and so, although I’d rather not be a PE, I might be able to help some anyway.

And regardless, best of luck with your paper!

(p.s. Yes, it is true that I served as founding editor of The Political Methodologist, but it was the “founding” part of “founding editor” that I enjoyed!)

Dominic Skinnion

Mon, 01 Jan 0001 00:00:00 +0000

Donald B. Rubin

Mon, 01 Jan 0001 00:00:00 +0000

Duncan J. Watts

Mon, 01 Jan 0001 00:00:00 +0000

Eleanor Powell

Mon, 01 Jan 0001 00:00:00 +0000

Electronic Collection Development in the Harvard College Library

Mon, 01 Jan 0001 00:00:00 +0000

Gary King, 1996. Download the report (PDF)

This document describes the organization, decision processes, and operating procedures by which the Harvard College Library (HCL) acquires electronic data. It first explains HCL’s long-established book collection procedures and then turns to electronic data acquisitions. The document concludes with some possible issues for discussion. The information herein was gathered from interviews with library staff, all of whom were exceedingly helpful.

Elizabeth Desombre

Mon, 01 Jan 0001 00:00:00 +0000

Elizabeth Kolster

Mon, 01 Jan 0001 00:00:00 +0000

Elizabeth Rosenblatt

Mon, 01 Jan 0001 00:00:00 +0000

Elizabeth Stuart

Mon, 01 Jan 0001 00:00:00 +0000

Emmanuela Gakidou

Mon, 01 Jan 0001 00:00:00 +0000

Eric Mazur

Mon, 01 Jan 0001 00:00:00 +0000

Eric Reinhardt

Mon, 01 Jan 0001 00:00:00 +0000

Ethan Katz

Mon, 01 Jan 0001 00:00:00 +0000

Etienne Krug

Mon, 01 Jan 0001 00:00:00 +0000

Evaluating Social Security Forecasts

Mon, 01 Jan 0001 00:00:00 +0000

The accuracy of U.S. Social Security Administration (SSA) demographic and financial forecasts is crucial for the solvency of its Trust Funds, government programs comprising greater than 50% of all federal government expenditures, industry decision making, and the evidence base of many scholarly articles. Forecasts are also essential for scoring policy proposals, put forward by both political parties. Because SSA makes public little replication information, and uses ad hoc, qualitative, and antiquated statistical forecasting methods, no one in or out of government has been able to produce fully independent alternative forecasts or policy scorings. Yet, no systematic evaluation of SSA forecasts has ever been published by SSA or anyone else. We show that SSA's forecasting errors were approximately unbiased until about 2000, but then began to grow quickly, with increasingly overconfident uncertainty intervals. Moreover, the errors all turn out to be in the same potentially dangerous direction, each making the Social Security Trust Funds look healthier than they actually are. We also discover the cause of these findings with evidence from a large number of interviews we conducted with participants at every level of the forecasting and policy processes. We show that SSA's forecasting procedures meet all the conditions the modern social-psychology and statistical literatures demonstrate make bias likely. When those conditions mixed with potent new political forces trying to change Social Security and influence the forecasts, SSA's actuaries hunkered down trying hard to insulate themselves from the intense political pressures. Unfortunately, this otherwise laudable resistance to undue influence, along with their ad hoc qualitative forecasting models, led them to also miss important changes in the input data such as retirees living longer lives, and drawing more benefits, than predicted by simple extrapolations. We explain that solving this problem involves using (a) removing human judgment where possible, by using formal statistical methods -- via the revolution in data science and big data; (b) instituting formal structural procedures when human judgment is required -- via the revolution in social psychological research; and (c) requiring transparency and data sharing to catch errors that slip through -- via the revolution in data sharing & replication.

Articles and Presentations

Konstantin Kashin, Gary King, and Samir Soneji. 2015. "Explaining Systematic Bias and Nontransparency in US Social Security Administration Forecasts". Political Analysis, 23, 3, Pp. 336-62.

Konstantin Kashin, Gary King, and Samir Soneji. 2015. "Systematic Bias and Nontransparency in US Social Security Administration Forecasts". Journal of Economic Perspectives, 29, 2, Pp. 239-58.

Frequently Asked Questions

Related Materials

Federico Girosi and Gary King. 2008. Demographic Forecasting. Princeton: Princeton University Press.

Konstantin Kashin, Gary King, and Samir Soneji. 2015. "Replication Data For: Explaining Systematic Bias and Nontransparency in U.S. Social Security Administration Forecasts.". doi:6:967llFHgiywsHWWp1cVg9A.

Publisher's Version

Publisher's Version

Konstantin Kashin, Gary King, and Samir Soneji. 2015. "Replication Data For: Systematic Bias and Nontransparency in U.S. Social Security Administration Forecasts.". doi:5:1oerGFXQ0Bu9bcMFU5/t2A.

Publisher's Version

Publisher's Version

Konstantin Kashin, Gary King, and Samir Soneji. 2016. "Scoring Social Security Proposals: Response from Kashin, King, and Soneji". Journal of Economic Perspectives, 30, 2, Pp. 245-48.

Samir Soneji and Gary King. 2012. "Statistical Security for Social Security". Demography, 49, 3, Pp. 1037-60.

Gary King and Samir Soneji. 2011. "The Future of Death in America". Demographic Research, 25, 1, Pp. 1-38.

Evan Chen

Mon, 01 Jan 0001 00:00:00 +0000

Evann Smith

Mon, 01 Jan 0001 00:00:00 +0000

Evrhet Milam

Mon, 01 Jan 0001 00:00:00 +0000

Federico Girosi

Mon, 01 Jan 0001 00:00:00 +0000

Feng Zhang

Mon, 01 Jan 0001 00:00:00 +0000

Frederick M. Hess

Mon, 01 Jan 0001 00:00:00 +0000

George Yean

Mon, 01 Jan 0001 00:00:00 +0000

Georgina 'Georgie' Evans

Mon, 01 Jan 0001 00:00:00 +0000

Gi-Heon Kwon

Mon, 01 Jan 0001 00:00:00 +0000

Giuseppe Porro

Mon, 01 Jan 0001 00:00:00 +0000

Glen McGee

Mon, 01 Jan 0001 00:00:00 +0000

Gov 2020: The Hidden Curriculum

Mon, 01 Jan 0001 00:00:00 +0000

Semester: Fall

Year offered: 2025

- Syllabus
- Consolidated View of entire class, or direct links:

- For Instructors (preview for students): Classes, Lectures, Sections
- For Students: Assignments, Canvas

Summary: Gov 2020 has two components:

A popular semester-long project where students learn how to write and publish an article in a scholarly journal, beginning with an article replication. This assignment, an early version of which was described in the article "Publication, Publication" and previously part of Gov 2001*, has resulted in many students' first publications, conference presentations, dissertations, and awards.
A collection of projects designed to show how science (and, by extension, the profession) works. It includes what a big idea is in our field; developing defensible answers; co-authoring; writing for impact; effective presentations; solving problems by changing the question; going beyond publication to impact; managing the transition academics undergo from private to public figure; leveraging universities, startups, industry, and government for research; and more.

A considerable body of wisdom about how to succeed in our profession accumulates informally as each PhD cohort learns, improves, and passes knowledge on to its successors. Most of this "hidden curriculum" is not taught in formal classes, but is essential to making the most of graduate school, producing high quality research, and securing the best jobs. Unfortunately, the 2020 pandemic severed the connection between cohorts, leaving individual students to try to piece together this knowledge on their own. We formalize much of this informal wisdom to provide 20/20 vision (remember the course number yet?) into the hidden curriculum and go well beyond what had been possible even before the pandemic, and increase the chances of success in graduate school and the profession.

The biggest impact methodologists make when asked for advice is rarely the answers to technical questions. Instead, it is providing the big picture and reorienting research directions. Many often pose what they think is a technical question (and sometimes sounds like "Someone told me I have a problem called `endogeneity'. Where is the button in SPSS to fix it?"), but which has no answer, isn't even a coherent question, or doesn't matter even if it could be answered. When we get them to tell us about their project, we may learn that they have unsolvable problems (such as impossible identification issues, 85% missing data, theory and data with little correspondence, or research questions that no one would care about even if they could be answered, etc.). When successful, we work out the bigger picture with them and reframe the research question into one that other scholars would care about and which can be answered without impossible-to-defend technical assumptions. Completion, fencing with anonymous reviewers, and publication become far easier too. This is by far the most common impact good methodologists have on those around them. There's no time to teach this in technical methods classes, but you will learn it in Gov 2020.

-----------
* This assignment was previously part of Gov2001. (A new version of Gov 2001, excluding this replication exercise, is now offered by others.) Materials from when I taught Gov 2001 remain available, including my online class lectures; the book Unifying Political Methodology; an app 2k1-in-silico: An Interactive Methods Non-Textbook and related paper Statistical Intuition Without Coding (or Teachers); Perusall.com, and the course website.

Grace Kim

Mon, 01 Jan 0001 00:00:00 +0000

Greg Adams

Mon, 01 Jan 0001 00:00:00 +0000

Gregory Wawro

Mon, 01 Jan 0001 00:00:00 +0000

Han Altae-Tran

Mon, 01 Jan 0001 00:00:00 +0000

Hannah Bayer

Mon, 01 Jan 0001 00:00:00 +0000

Hanspeter Pfister

Mon, 01 Jan 0001 00:00:00 +0000

Heather K. Gerken

Mon, 01 Jan 0001 00:00:00 +0000

Heather Stoll

Mon, 01 Jan 0001 00:00:00 +0000

Héctor Hernández Llama

Mon, 01 Jan 0001 00:00:00 +0000

Helen Bentley

Mon, 01 Jan 0001 00:00:00 +0000

Helen Margetts

Mon, 01 Jan 0001 00:00:00 +0000

Hendrik Strobelt

Mon, 01 Jan 0001 00:00:00 +0000

How to Write a Publishable Paper

Mon, 01 Jan 0001 00:00:00 +0000

This site shows how to write a publishable article by beginning with the replication of a previously published article. Following the advice has long been a central assignment for Quantitative Social Science Methods, I, and recently has been moved to Gov2020: The Hidden Curriculum. After fine tuning these suggestions over many years, I published the 2006 version as:

Gary King Publication, Publication, PS: Political Science and Politics, Vol. XXXIX, No. 1 (January, 2006), 119-125 (Abstract: HTML | Article: PDF)

Continuing updates to this article can be found here:

When hunting for data associated with an article, carefully read all footnotes, appendices, tables, captions, web appendices, etc. Also, check previous or subsequent papers from the same author on the same or related subjects for better documentation.
Before you contact an author, check his or her website, Dataverse, the ICPSR Publications Related Archive, and the journal's web site to see you can find the replication materials on your own. If you need to contact the author of the original article, consolidate your requests into as few emails as possible; no matter how generous they are with their time, they will be the bottleneck and you may need to contact them again when trying to understand the data.
To increase the probability that your paper will eventually be published, its usually better to choose an article from a better journal. The way these things work, if you find something important and do a good job researching and writing your paper, you will have a chance at publication in that journal. If your submission is rejected for whatever reason, odds are you have to go down one level in the hierarchy of journals. It's thus best to start with some of the best journals. Of course, some terrific -- and influential -- articles are not in the most visible journals, so this is a consideration but not a rule.
Although throughout the process you should do whatever you can to avoid a conflictual relationship with the authors of the article you replicate, for class eliminate the possibility altogether: Don't start wtih those with whom you have ongoing professional relationships, such as faculty on your committee or in your department. Since there are so many other options out there, its easy to avoid even potential conflicts like these by making wise choices now. This will also help you ensure during the research that your decision making is not influenced by anything but producing the best work.
This is probably the first paper you are writing that is not about you: Explaining how hard you worked, that you included everything the professor asked for, and how smart you are, are now all irrelevant distractions. Your goal instead is to construct a paper that makes people want to read it. Your title, abstract, and paper each must focus on what others will learn if they devote the time to read it. You must answer "Whose mind are you going to change about what?", such as by starting the abstract with "In this paper, we demonstrate that...".
Additional points just for the project in my class:
1. You are required to ask for my advice on the article you are considering replicating (whether you take the advice is up to you!). Please stop by my office with your coauthors and copies of the article you are considering (no, you don't need an appointment), or send me a PDF and CC your coauthor. (Please don't forget: the article should be published within the last few years, from a good journal, and use methods we have or will talk about in class, or at about the same level of sophistication.)
2. Since we will give the first draft of your replication, data, and code to another student in the class to replicate your replication, you must use R for this part of the paper (it wouldn't be fair to ask another student to learn new software that our TFs don't support just for that purpose). For other parts of your work, or for extending it, you're welcome to use whatever tools you desire. Remember, you also don't need to replicate every part of the article you choose; just the part you (and not necessarily the author of the original article) can justify as important for the paper you will write from this.
3. Choose an article with data that you are allowed to share publicly, without any restriction. You not only need permission to use the data, but also to share it with others in the class and beyond.
4. Your paper must use some methods at least as advanced as those we learned in this class; that means that if you choose an article with less advanced methods (such as only linear regression), your paper will only work as a class project if you have a more advanced method that makes sense to use and if it produces sufficiently worthwhile results that justifies itself. Since introducing a new method into a paper when it doesn't make a difference doesn't make for a good paper, you are at more risk for the class project if you choose an article that uses relatively simple statistical procedures. Its not necessarily the wrong choice, since if the author is using simple procedures and you have better ones, you might be able to extract more information.
5. Choose a reasonable sample size: (i) Do not choose an article with too massive a data set. Larger data sets can of course be more informative, but if they overwhelm the computational resources you have available you may need to spend a disproportionate amount of your time overcoming these problems. (ii) Do not choose an article with too "uninformative" (which usually but not always means too "small") a sample size.
6. After you have your results and before you start to write the paper, prepare an abstract of 150 words or less and post it in the appropriate Perusall channel. We will all comment on it and try to help you improve it, and thereby the paper. After you've finished the analysis, you have borne most of the costs of the research project, and so it is at precisely this time when you can sometimes most easily have a big impact on improving the final product.
7. Prepare the paper double-spaced with at least 1 inch margins all around and in 12 point font. (I realize that you can see it in smaller fonts, but that's not necessarily true for your reviewers.) Overall, make the style of the paper look like those professors write. For examples, see my preprints.
8. Do not choose an article unless you fully understand its argument, methods, theory, and substance.
9. Please read "Publication, Publication" and this update carefully and check it repeatedly. Please try to avoid us having to refer you back to this material when we give you final comments on your paper.

Suggestions for other instructors who might wish to use this assignment in your classes. The key to remember is that students are rarely good at coming up with a big publishable idea, or sometimes even an answerable question, on their own. They will eventually be good at this, but this is essentially their first real effort. To increase the probability that this experience will be a success:

Meet with them collectively or in small groups to help them construct their arguments, decide what avenues to pursue, and construct a winning argument. Its ok for you (or your teaching assistants) to give them the key idea for their paper if they're doing the hard work of replication and analysis.
Keep them focused on satisfying each and every item on the list. Encourage them to read it over multiple times while they are preparing, while they are writing, and before they turn it in.
Before turning in anything, require them to have another student verify that they meet each item in the checklist in "Publication, Publication." Missing these is way too common and much easier to see in someone else's work than your own.
Use the exchange of abstracts (preferably in a way so that everyone else can see at the same time) as a way to help them with their overall pitch, the organization of their paper, and its main point.
You can't emphasize enough how rigorously organized and concise their paper must be. The section headings alone ought to convey the entire message of the paper. Same for the title, for the abstract, and for the introduction.

Hung-Der Fu

Mon, 01 Jan 0001 00:00:00 +0000

Ian Ayres

Mon, 01 Jan 0001 00:00:00 +0000

Introduction to Quantitative Political Methodology, G1000

Mon, 01 Jan 0001 00:00:00 +0000

Year offered: 2010

Not offered by me this year. Introduction to major quantitative technqiues used in political science. Covers exploratory data analysis, as well as descriptive and causal statistical inference of many types. The course emphasizes probability theory, regression analysis and other statistical techniques, and uses new techniques of stochastic simulation to get answers easily and to interpret statistical results in a manner very close to the political substance of the problem at hand. NOTE: Frequently taken by undergraduates needing quantitative techniques for thesis research and by graduate students satisfying department requirements. This course also serves as the first in a series of three quantitative courses offered by the department.

Jack Deschler

Mon, 01 Jan 0001 00:00:00 +0000

James Alt

Mon, 01 Jan 0001 00:00:00 +0000

James Briggs

Mon, 01 Jan 0001 00:00:00 +0000

James Fowler

Mon, 01 Jan 0001 00:00:00 +0000

James Honaker

Mon, 01 Jan 0001 00:00:00 +0000

James McCann

Mon, 01 Jan 0001 00:00:00 +0000

James Tompkin

Mon, 01 Jan 0001 00:00:00 +0000

Jasjeet Sekhon

Mon, 01 Jan 0001 00:00:00 +0000

Jason Lakin

Mon, 01 Jan 0001 00:00:00 +0000

Jason Wittenberg

Mon, 01 Jan 0001 00:00:00 +0000

Jason Wittenburg

Mon, 01 Jan 0001 00:00:00 +0000

Jean Aurambault

Mon, 01 Jan 0001 00:00:00 +0000

Jeff Gill

Mon, 01 Jan 0001 00:00:00 +0000

Jeff Lewis

Mon, 01 Jan 0001 00:00:00 +0000

Jeffrey C. Blossom

Mon, 01 Jan 0001 00:00:00 +0000

Jeffrey Segal

Mon, 01 Jan 0001 00:00:00 +0000

Jennifer Hill

Mon, 01 Jan 0001 00:00:00 +0000

Jennifer Pan

Mon, 01 Jan 0001 00:00:00 +0000

Joel Newberger

Mon, 01 Jan 0001 00:00:00 +0000

John Boscardin

Mon, 01 Jan 0001 00:00:00 +0000

John M. Bruce

Mon, 01 Jan 0001 00:00:00 +0000

John Openshaw

Mon, 01 Jan 0001 00:00:00 +0000

Jonathan Bischof

Mon, 01 Jan 0001 00:00:00 +0000

Jonathan M. Spector

Mon, 01 Jan 0001 00:00:00 +0000

Jonathan N. Katz

Mon, 01 Jan 0001 00:00:00 +0000

Jonathan Ullman

Mon, 01 Jan 0001 00:00:00 +0000

Jonathan Wand

Mon, 01 Jan 0001 00:00:00 +0000

Josephine T. Andrews

Mon, 01 Jan 0001 00:00:00 +0000

Josh Inkenbrandt

Mon, 01 Jan 0001 00:00:00 +0000

Joshua Salomon

Mon, 01 Jan 0001 00:00:00 +0000

Joshua Tucker

Mon, 01 Jan 0001 00:00:00 +0000

Juan Eugenio Hernández Ávila

Mon, 01 Jan 0001 00:00:00 +0000

Juan M. Lavista Ferres

Mon, 01 Jan 0001 00:00:00 +0000

Justin Grimmer

Mon, 01 Jan 0001 00:00:00 +0000

Assistant Professor in the Department of Political Science at Stanford University. His research areas include political methodology and political behavior. His work focuses on estimating the causal effects of social networks and political/social elites on mass political behavior. He has also implemented ARIMA time series models into Zelig and is currently developing and applying unsupervised learning routines for a variety of applications in political science.

Karen Ferree

Mon, 01 Jan 0001 00:00:00 +0000

Katherine Clayton

Mon, 01 Jan 0001 00:00:00 +0000

Katherine Irajpanah

Mon, 01 Jan 0001 00:00:00 +0000

Katherine Semrau

Mon, 01 Jan 0001 00:00:00 +0000

Katie Colton

Mon, 01 Jan 0001 00:00:00 +0000

Kay Schlozman

Mon, 01 Jan 0001 00:00:00 +0000

Kelly Miller

Mon, 01 Jan 0001 00:00:00 +0000

Kenesia Washington

Mon, 01 Jan 0001 00:00:00 +0000

Kenji Shibuya

Mon, 01 Jan 0001 00:00:00 +0000

Kenneth Benoit

Mon, 01 Jan 0001 00:00:00 +0000

Kenneth Scheve

Mon, 01 Jan 0001 00:00:00 +0000

Kenneth Shotts

Mon, 01 Jan 0001 00:00:00 +0000

Kevin Quinn

Mon, 01 Jan 0001 00:00:00 +0000

Kobbi Nissim

Mon, 01 Jan 0001 00:00:00 +0000

Konstantin Kashin

Mon, 01 Jan 0001 00:00:00 +0000

Kosuke Imai

Mon, 01 Jan 0001 00:00:00 +0000

Kristian Gleditsch

Mon, 01 Jan 0001 00:00:00 +0000

Lada Adamic

Mon, 01 Jan 0001 00:00:00 +0000

Langche Zeng

Mon, 01 Jan 0001 00:00:00 +0000

Larry J. Sabato

Mon, 01 Jan 0001 00:00:00 +0000

Latanya Sweeney

Mon, 01 Jan 0001 00:00:00 +0000

Learning from College, after College: A Commencement Speech at SUNY New Paltz

Mon, 01 Jan 0001 00:00:00 +0000

From the 182nd May commencement of the State University of New York at New Paltz, where Gary received an honorary degree on 23 May 2010.

Download: Gary’s speech (PDF) · Honorary degree citation, by the New Paltz President (PDF)

Learning from College, after College

This is by far and away my second favorite degree!

Thanks so much for the introduction. But just so you know, when those of us from here want to pay Harvard a compliment, we call it “the New Paltz of the Northeast.”

First of all, a big congratulations to the Classes of 2009 and 2010! Let’s also thank those who made all this possible. Let’s hear it for your moms and dads… your brothers and sisters… your aunts and uncles… your friends and classmates… and the New Paltz faculty and staff. Not much would have happened without you all.

Well here you are, in the top few percent of the world’s population with a college degree. Today marks an awesome personal accomplishment — years of your life devoted to education, art, and science — incredibly important for your future, but as important for what you can now do for the rest of society. So this is one small party for you, and one giant leap for mankind.

OK, but enough about you…. Enough about how great you are; mom and dad will take that up again later.

My job is to help you look to the next challenge. And I come with only one message, one that will help you more than any other I can think of: what you learn from New Paltz doesn’t end today; as you go forward, think back to what you experienced here, and this place will become a fountain of lessons and strength to draw upon for the rest of your life.

You’ll learn, as I did, how to translate your New Paltz experiences into lessons, discoveries, insights, successes, careers, families, and perspectives. What New Paltz gives you doesn’t stop with a degree. I’ll give some examples of this from my experience here.

I learned here about government, politics, communications, astronomy, biology, and statistics. I learned how to write, how to think analytically, how science works, and how to program computers. I also learned how to stay in the library and work for more than 20 minutes at a time. I eventually figured out, as you all did, how to get passing grades.

And by the way, I hear a few of you actually didn’t get straight As. Now, dear parents, this was not because your kids partied too much; it’s because budget cuts caused the Registrar’s office to make lots of transcription errors. Don’t complain to your kids; help out the faculty and staff and write to the state legislature! In fact I’m pretty sure that’s what happened in one class I took that will remain nameless; isn’t that right, Professor Brownstein?

Fortunately, I’m authorized to let you in on the big secret: Grades don’t matter any more! You’ve graduated; and no one can take away your degree or what you learned here.

Now I’m sure you learned a ton in your classes, but don’t forget what you learned outside of class. I learned about the value of friends and made friends I’ve had my whole life. Some are even crazy enough to be here today. I hope my daughter, who is also here, and whom I am so proud of by the way, is as lucky when she goes to college. I just wish that my friends don’t tell my daughter all the things that went on here!

In any event, some of what I learned in New Paltz is that the advice commencement speakers give 22-year-olds is just nuts.

For example, Winston Churchill told some graduates: “Never give in, never give in, never, never, never never — in nothing great or small, large or petty — never give in.” Well, if you’re fighting a just war, that sounds like great advice. But can you imagine living with this guy? Sure, pursue your dreams; don’t let anyone get in your way; and by all means win. But you’re allowed to learn from those around you too. You’re allowed to use judgment and decide to go in a new direction.

I remember when I learned this lesson. As a Freshman, I lived on the first floor of Scudder Hall. At 3am one morning, noise coming from the hallway woke me up, and I was furious. They were breaking all the rules. I strutted out to the hallway to find an epic water fight, 2 inches of water on the floor, and a friend with a spaghetti pot on his head. After 2 minutes of being indignant, I got drenched and then tackled by my friends who decided not to take me as seriously as I was taking myself. Eventually, I realized no one was getting hurt, everyone was having fun but me, and although by the rules I was completely right and they were totally wrong, this wasn’t a war. In those 2 minutes, I saw most of what I now know about the advantage of seeing the world from the perspective of others.

So I’ll admit I spent the next 2 hours spraying water everywhere and sliding down the corridor with everyone else. And I learned: Sometimes you stand your ground and never give in; sometimes you learn what others see differently and change your mind; and sometimes you put a spaghetti pot on your head and slide down the hallway. Whenever I have a disagreement back in Cambridge, I remember that water fight.

Let me tell you a different story. Lots of commencement speakers tell the graduates to take big bold risks. This is bizarre because commencement speakers are older and research shows that older folks don’t take big risks; it’s the young who are too often taking bigger risks than they should. I think commencement speakers are confused because you don’t get to be a commencement speaker unless you’re lucky a lot of times in a row!

I was an RA in New Paltz, and I remember driving up to the dorm and seeing a friend sitting on his 3rd floor windowsill. He had a rope tied around the bar in his closet and then through his belt and around his back. Like a rock climber, he was planning to rappel down the outside of the building from the 3rd floor!

I admit, it looked like a load of fun, and I still want to try it, but I managed to stop him without killing himself and without killing me. From this experience and others, I learned about managing people, and taking calculated rather than crazy risks; I didn’t know the importance of this event until later, but it’s been helpful to think about many times.

So go out there and do some great things, and don’t delay. Remember what my grandmother said: “one minute you’re 16, the next thing you know you’re 87.” So use the time; go do some cool stuff; and come back and teach us all.

But whatever you do, remember what you learned here, remember this place, remember the people, and continually learn from your experiences here. You can forget the commencement speaker (and you will!), but don’t forget New Paltz.

Thanks so much for the honor of letting me share this day with you.

More SUNY New Paltz photos (by Danny Wild)

Lee Epstein

Mon, 01 Jan 0001 00:00:00 +0000

Leonid Andreev

Mon, 01 Jan 0001 00:00:00 +0000

Letters of Recommendation

Mon, 01 Jan 0001 00:00:00 +0000

If you need a letter of recommendation from a faculty member, and we've worked closely together in some capacity beyond class (e.g., RA, TF, co-author, advisee, etc.), I'm happy to help out. However, if you have worked more closely with some other faculty member, a letter from that person is much more likely to be useful to you. Wherever you are applying will not merely want to know that someone endorses your candidacy (every letter does that!); the crucial component is that the person writing the letter knows you well, through a diverse array of experiences together.

If you think that person is me, then please send me, along with your request, one email to king@harvard.edu with the following as separate (nonzipped) attachments in PDF format (not Word and no paper please): copies of whatever documents you plan to submit to the organization to which you are applying that you think will be useful; a curriculum vitae; transcripts; test scores, etc.; and a confidential "sliced bread" memo (i.e., why I am better than...) addressed to me. This memo should not be a draft letter of recommendation, but instead should include a list of bullet points you want to make sure I remember when I start writing. These items can include how I got to know you and standard cv items (prizes won, etc.), but should also include relevant anecdotes that might make useful stories to illustrate characteristics of you or your work (e.g., solved a problem in five minutes that King had been working on full time for six weeks). This is not the time for modesty (only you and I will ever see it), and don't assume I will remember something if you exclude it. Include a very short (2-3 sentence) abstract or main finding of your thesis or other current work. If your record includes something negative (such as if you were in graduate school for too long), include your plausible explanation for what happened, if there is one.

Please use my assistant's email king-assist@iq.harvard.edu, instead of mine, if the organization to which you are applying uses a web interface, it requests a place to send automated emails, or you use a dossier service. This will also enable us to make certain your letter goes out where and when it should.

As much lead time as you can provide would be appreciated.

Libby Jenke

Mon, 01 Jan 0001 00:00:00 +0000

Lisa Martin

Mon, 01 Jan 0001 00:00:00 +0000

Lisa R. Hirschhorn

Mon, 01 Jan 0001 00:00:00 +0000

Lyn Ragsdale

Mon, 01 Jan 0001 00:00:00 +0000

Manett Vargas

Mon, 01 Jan 0001 00:00:00 +0000

Marco Gaboardi

Mon, 01 Jan 0001 00:00:00 +0000

Margaret E. 'Molly' Roberts

Mon, 01 Jan 0001 00:00:00 +0000

Margaret Trevor

Mon, 01 Jan 0001 00:00:00 +0000

Mark A. Travassos

Mon, 01 Jan 0001 00:00:00 +0000

Mark Diggory

Mon, 01 Jan 0001 00:00:00 +0000

Markus Strohmaier

Mon, 01 Jan 0001 00:00:00 +0000

Marshall Van Alstyne

Mon, 01 Jan 0001 00:00:00 +0000

Martha María Téllez-Rojo

Mon, 01 Jan 0001 00:00:00 +0000

Martin Tanner

Mon, 01 Jan 0001 00:00:00 +0000

Math Prefresher for Political Scientists (Faculty advisor)

Mon, 01 Jan 0001 00:00:00 +0000

Year offered: 2023

Not only do the quantitative and formal modeling courses at Harvard require mathematics and computer programming | it's becoming increasingly difficult to take courses in political economy, American politics, comparative politics, or international relations without encountering game-theoretic models or statistical analyses. One need only flip through the latest issues of the top political science journals to see that mathematics have entered the mainstream of political science. Unfortunately, most undergraduate political science programs have not kept up with this trend, and first-year graduate students often find themselves lacking in basic technical skills. This course is not intended to be an introduction to game theory or quantitative methods. Rather, it introduces basic mathematics and computer skills needed for quantitative and formal modeling courses offered at Harvard. No tests are administered or grades are given out, but almost all incoming graduate students in Government take the course. The course is offered every year in the two weeks before classes start. See the Math Prefresher Web Site.

Matthew Blackwell

Mon, 01 Jan 0001 00:00:00 +0000

Graduate student in the Department of Government at Harvard. He likes statistical methodology and game theory, but cannot seem to find the perfect place to apply them. He co-wrote the Amelia II package and frontend for R.

Matthew J. Salganik

Mon, 01 Jan 0001 00:00:00 +0000

Matthew Knowles

Mon, 01 Jan 0001 00:00:00 +0000

Matthew Lebo

Mon, 01 Jan 0001 00:00:00 +0000

Mauricio Hernández Ávila

Mon, 01 Jan 0001 00:00:00 +0000

Max Goplerud

Mon, 01 Jan 0001 00:00:00 +0000

Maxwell Palmer

Mon, 01 Jan 0001 00:00:00 +0000

Maya Sen

Mon, 01 Jan 0001 00:00:00 +0000

Mayya Komisarchik

Mon, 01 Jan 0001 00:00:00 +0000

Meg Schwenzfeier

Mon, 01 Jan 0001 00:00:00 +0000

Melissa Sands

Mon, 01 Jan 0001 00:00:00 +0000

Mercè Crosas

Mon, 01 Jan 0001 00:00:00 +0000

Meredith Dost

Mon, 01 Jan 0001 00:00:00 +0000

Micah Altman

Mon, 01 Jan 0001 00:00:00 +0000

Michael Croatto

Mon, 01 Jan 0001 00:00:00 +0000

Michael Gill

Mon, 01 Jan 0001 00:00:00 +0000

Michael Gilligan

Mon, 01 Jan 0001 00:00:00 +0000

Michael Herron

Mon, 01 Jan 0001 00:00:00 +0000

Michael Krot

Mon, 01 Jan 0001 00:00:00 +0000

Michael Laver

Mon, 01 Jan 0001 00:00:00 +0000

Michael Macy

Mon, 01 Jan 0001 00:00:00 +0000

Michael McDonald

Mon, 01 Jan 0001 00:00:00 +0000

Michael New

Mon, 01 Jan 0001 00:00:00 +0000

Michael Ting

Mon, 01 Jan 0001 00:00:00 +0000

Michael Tomz

Mon, 01 Jan 0001 00:00:00 +0000

Michail Schwab

Mon, 01 Jan 0001 00:00:00 +0000

Miguel Solano

Mon, 01 Jan 0001 00:00:00 +0000

Mireille Kamariza

Mon, 01 Jan 0001 00:00:00 +0000

Miscellanea

Mon, 01 Jan 0001 00:00:00 +0000

Commercial Startups: Crimson Hexagon (merged with Brandwatch), Learning Catalytics (acquired by Pearson), Perusall, Thresher, OpenScholar.
Social Science One: Building Industry-Academic Partnerships
“Gary King” in Chinese. Mainland: 金加里, Taiwan: 金·蓋瑞
Do you have a paper to submit to PNAS?
Are you applying to our Ph.D. program at Harvard?
“Learning from College, after College: A Commencement Speech at SUNY New Paltz” — From The 182nd May Commencement Of The State University Of New York At New Paltz.
National Academy of Sciences “Registry of Members” (photo); “QnAs with Gary King” Interview by PNAS (PDF).
Electronic Collection Development in the Harvard College Library (Gary King, 1996)

Mitchell Brooks

Mon, 01 Jan 0001 00:00:00 +0000

Mitsuru Mukaigawara

Mon, 01 Jan 0001 00:00:00 +0000

Miya Woolfalk

Mon, 01 Jan 0001 00:00:00 +0000

Miyoung Chun

Mon, 01 Jan 0001 00:00:00 +0000

Mohan (Penubarti) Rao

Mon, 01 Jan 0001 00:00:00 +0000

Myron Gutmann

Mon, 01 Jan 0001 00:00:00 +0000

Nancy Billica

Mon, 01 Jan 0001 00:00:00 +0000

Nancy Burns

Mon, 01 Jan 0001 00:00:00 +0000

Narender Sharma

Mon, 01 Jan 0001 00:00:00 +0000

Natalie Ayers

Mon, 01 Jan 0001 00:00:00 +0000

Nathaniel Beck

Mon, 01 Jan 0001 00:00:00 +0000

Nathaniel Persily

Mon, 01 Jan 0001 00:00:00 +0000

Naunihal Singh

Mon, 01 Jan 0001 00:00:00 +0000

Neelam Dhingra-Kumar

Mon, 01 Jan 0001 00:00:00 +0000

Nicholas Christakis

Mon, 01 Jan 0001 00:00:00 +0000

Nicole Nova

Mon, 01 Jan 0001 00:00:00 +0000

Niels Tomijima

Mon, 01 Jan 0001 00:00:00 +0000

Nirmala Ravishankar

Mon, 01 Jan 0001 00:00:00 +0000

Norman Nie

Mon, 01 Jan 0001 00:00:00 +0000

Noshir Contractor

Mon, 01 Jan 0001 00:00:00 +0000

Olivia Lau

Mon, 01 Jan 0001 00:00:00 +0000

Ophir Shalem

Mon, 01 Jan 0001 00:00:00 +0000

Ori Rosen

Mon, 01 Jan 0001 00:00:00 +0000

Pamela Ban

Mon, 01 Jan 0001 00:00:00 +0000

Patrick Brandt

Mon, 01 Jan 0001 00:00:00 +0000

Patrick Lam

Mon, 01 Jan 0001 00:00:00 +0000

Patrick Wolf

Mon, 01 Jan 0001 00:00:00 +0000

Patrik Gothe

Mon, 01 Jan 0001 00:00:00 +0000

Paul Brace

Mon, 01 Jan 0001 00:00:00 +0000

Paul W. Glimcher

Mon, 01 Jan 0001 00:00:00 +0000

People

Mon, 01 Jan 0001 00:00:00 +0000

Peter Dyrud

Mon, 01 Jan 0001 00:00:00 +0000

Philip Paolino

Mon, 01 Jan 0001 00:00:00 +0000

Prateek Goorha

Mon, 01 Jan 0001 00:00:00 +0000

Quantitative Social Science Methods, I: Government 2001, and E-200

Mon, 01 Jan 0001 00:00:00 +0000

Semester: Fall

Year offered: 2022

Gov2001 is the first (or sometimes second) course for incoming Harvard Government Department PhD students; also taken by graduate and undergraduate students in other departments, and students elsewhere through the Harvard Extension School as Stat E-200. For details, see the class website: j.mp/G2001, where you can also access all the lecture videos via YouTube or, with collaborative annotation via Perusall.com; a link to a no-coding-required AI assistant; my book Unifying Political Methodology; and other class materials.

Rainer Winkelmann

Mon, 01 Jan 0001 00:00:00 +0000

Rakesh Sarwal

Mon, 01 Jan 0001 00:00:00 +0000

Rebecca Firestone

Mon, 01 Jan 0001 00:00:00 +0000

Research Areas

Mon, 01 Jan 0001 00:00:00 +0000

Richard A. Berk

Mon, 01 Jan 0001 00:00:00 +0000

Richard A. Nielsen

Mon, 01 Jan 0001 00:00:00 +0000

Richard Johnston

Mon, 01 Jan 0001 00:00:00 +0000

Richard Merelman

Mon, 01 Jan 0001 00:00:00 +0000

Richard R.W. Brooks

Mon, 01 Jan 0001 00:00:00 +0000

Richard Tucker

Mon, 01 Jan 0001 00:00:00 +0000

Rifat Atun

Mon, 01 Jan 0001 00:00:00 +0000

Robert Murray

Mon, 01 Jan 0001 00:00:00 +0000

Robert O. Keohane

Mon, 01 Jan 0001 00:00:00 +0000

Robert Walker

Mon, 01 Jan 0001 00:00:00 +0000

Robert X Browning

Mon, 01 Jan 0001 00:00:00 +0000

Rockli Kim

Mon, 01 Jan 0001 00:00:00 +0000

Rumya Raghavan

Mon, 01 Jan 0001 00:00:00 +0000

Russ Mayer

Mon, 01 Jan 0001 00:00:00 +0000

Ruth Greenwood

Mon, 01 Jan 0001 00:00:00 +0000

Ryan Kennedy

Mon, 01 Jan 0001 00:00:00 +0000

Ryan Moore

Mon, 01 Jan 0001 00:00:00 +0000

Ryan Probasco

Mon, 01 Jan 0001 00:00:00 +0000

S. V. Subramanian

Mon, 01 Jan 0001 00:00:00 +0000

Salil Vadhan

Mon, 01 Jan 0001 00:00:00 +0000

Samir Soneji

Mon, 01 Jan 0001 00:00:00 +0000

Samuel S.-H. Wang

Mon, 01 Jan 0001 00:00:00 +0000

Sandra González-Bailón

Mon, 01 Jan 0001 00:00:00 +0000

Sascha Riaz

Mon, 01 Jan 0001 00:00:00 +0000

Scott Desposato

Mon, 01 Jan 0001 00:00:00 +0000

Scott Yockel

Mon, 01 Jan 0001 00:00:00 +0000

Sean Carey

Mon, 01 Jan 0001 00:00:00 +0000

Sebastian Bauhoff

Mon, 01 Jan 0001 00:00:00 +0000

Shannon L. Parker

Mon, 01 Jan 0001 00:00:00 +0000

Sherod Thaxton

Mon, 01 Jan 0001 00:00:00 +0000

Shiro Kuriwaki

Mon, 01 Jan 0001 00:00:00 +0000

Should You Use AI to Write for You?

Mon, 01 Jan 0001 00:00:00 +0000

Your goal is to produce the best work you are capable of. If AI helps you do that, great.

Also be aware that the many highly talented people around you have high expectations for you and your work product, whereas LLMs are trained to produce text at the level of the average document on the web. Ask yourself whether you can do better, and understand that others will inevitably judge you by the work you put your name on.

(Harvard's AI policy for students has no content other than that individual faculty must adopt a policy and convey it to their classes. This is my student AI policy, but perhaps some others will also consider this as relevant advice.)

Shusei Eshima

Mon, 01 Jan 0001 00:00:00 +0000

Sidney Verba

Mon, 01 Jan 0001 00:00:00 +0000

Sinan Aral

Mon, 01 Jan 0001 00:00:00 +0000

Skyler Cranmer

Mon, 01 Jan 0001 00:00:00 +0000

Soledad Prillaman

Mon, 01 Jan 0001 00:00:00 +0000

Sophie E. Hill

Mon, 01 Jan 0001 00:00:00 +0000

Soubhik Barari

Mon, 01 Jan 0001 00:00:00 +0000

Stefan Wojcik

Mon, 01 Jan 0001 00:00:00 +0000

Stefano Iacus

Mon, 01 Jan 0001 00:00:00 +0000

Stephen Ansolabehere

Mon, 01 Jan 0001 00:00:00 +0000

Stephen Pettigrew

Mon, 01 Jan 0001 00:00:00 +0000

Steven D. Voss

Mon, 01 Jan 0001 00:00:00 +0000

Steven Koonin

Mon, 01 Jan 0001 00:00:00 +0000

Strategies of Political Inquiry, Government 2010

Mon, 01 Jan 0001 00:00:00 +0000

Year offered: 2010

Gary King, Robert Putnam, and Sidney Verba: not offered this academic year. If you could learn only one thing in graduate school, it should be how to do scholarly research. You should be able to assess the state of a scholarly literature, identify interesting questions, formulate strategies for answering them, have the methodological tools with which to conduct the research, and understand how to write up the results so they can be published. Although many graduate level courses address these issues of research design indirectly, we provide an explicit analysis of each. We take empirical evidence to be historical, quantitative, or anthropological and focus on the theory of descriptive and causal inference underlying both quantitative and qualitative research. Primarily for graduate students and undergraduates preparing for senior thesis research.

Attachments

Strategies of Political Inquiry, Government 2010 Syllabus

Stu Snydman

Mon, 01 Jan 0001 00:00:00 +0000

Stuart R. Lipsitz

Mon, 01 Jan 0001 00:00:00 +0000

Sue J. Goldie

Mon, 01 Jan 0001 00:00:00 +0000

Sundeep Iyer

Mon, 01 Jan 0001 00:00:00 +0000

Susan Athey

Mon, 01 Jan 0001 00:00:00 +0000

Suzanna De Boef

Mon, 01 Jan 0001 00:00:00 +0000

Tami Kim

Mon, 01 Jan 0001 00:00:00 +0000

Thorben Primke

Mon, 01 Jan 0001 00:00:00 +0000

Tim King

Mon, 01 Jan 0001 00:00:00 +0000

Timothy D. Wilson

Mon, 01 Jan 0001 00:00:00 +0000

Timothy Prinz

Mon, 01 Jan 0001 00:00:00 +0000

Tony Jebara

Mon, 01 Jan 0001 00:00:00 +0000

Tuan Huynh

Mon, 01 Jan 0001 00:00:00 +0000

Vinay Pratap Singh

Mon, 01 Jan 0001 00:00:00 +0000

Viridiana Rios

Mon, 01 Jan 0001 00:00:00 +0000

Vishwajeet Kumar

Mon, 01 Jan 0001 00:00:00 +0000

Weijie Li

Mon, 01 Jan 0001 00:00:00 +0000

Wendy Lu

Mon, 01 Jan 0001 00:00:00 +0000

Wenxin Jiang

Mon, 01 Jan 0001 00:00:00 +0000

Will Lowe

Mon, 01 Jan 0001 00:00:00 +0000

William E. Allen

Mon, 01 Jan 0001 00:00:00 +0000

William K. Barnett

Mon, 01 Jan 0001 00:00:00 +0000

Workshop in Applied Statistics

Mon, 01 Jan 0001 00:00:00 +0000

Year offered: 2017

The Applied Statistics Workshop (Gov 3009) meets all academic year, Wednesdays, 12pm-1:30pm, in CGIS K354. This workshop is a forum for faculty, graduate students, visiting scholars, and others in the area to present and discuss methodological or empirical work in progress in an interdisciplinary setting. The workshop features a tour of Harvard's statistical innovations and applications with weekly stops in different disciplines. Free lunch is provided. It is a great way to meet the community of data scientists across campus.

Xihong Lin

Mon, 01 Jan 0001 00:00:00 +0000

Xin Jin

Mon, 01 Jan 0001 00:00:00 +0000

Xiuyuan Zhang

Mon, 01 Jan 0001 00:00:00 +0000

Yi William Wei

Mon, 01 Jan 0001 00:00:00 +0000

Yichen Guan

Mon, 01 Jan 0001 00:00:00 +0000

Ying Lu

Mon, 01 Jan 0001 00:00:00 +0000

Yon Soo Park

Mon, 01 Jan 0001 00:00:00 +0000

Yun Xu

Mon, 01 Jan 0001 00:00:00 +0000

Yusaku Horiuchi

Mon, 01 Jan 0001 00:00:00 +0000

Zachary J. Ward

Mon, 01 Jan 0001 00:00:00 +0000

Zagreb Mukerjee

Mon, 01 Jan 0001 00:00:00 +0000

Zheng Ma

Mon, 01 Jan 0001 00:00:00 +0000