HOME /

Presentations

Statistically Valid Inferences from Privacy Protected Data (MIT Analytics Lab), at MIT Analytics Lab, Sloan School (Building E62, Room 276), Thursday, November 3, 2022:
Slides

Venerable procedures for privacy protection and data sharing within academia, companies, and governments, and between sectors, have been proven to be completely inadequate (e.g., respondents in de-identified surveys can usually be re-identified). At the same time, unprecedented quantities of data that could help social scientists understand and ameliorate the challenges of human society are presently locked away inside companies, governments, and other organizations, in part because of worries about privacy violations. We address these problems with a general-...

Read more about Statistically Valid Inferences from Privacy Protected Data (MIT Analytics Lab)
Statistically Valid Inferences from Privacy Protected Data (Princeton University), at Quantitative Social Science Colloquium, Princeton University, Friday, October 7, 2022:
Slides

Venerable procedures for privacy protection and data sharing within academia, companies, and governments, and between sectors, have been proven to be completely inadequate (e.g., respondents in de-identified surveys can usually be re-identified). At the same time, unprecedented quantities of data that could help social scientists understand and ameliorate the challenges of human society are presently locked away inside companies, governments, and other organizations, in part because of worries about privacy violations. We address these problems with a general-...

Read more about Statistically Valid Inferences from Privacy Protected Data (Princeton University)
Statistically Valid Inferences from Privacy Protected Data (University of Wisconsin), at Models, Experiments, and Data workshop (MEAD) at the University of Wisconsin-Madison, Wednesday, October 5, 2022:
Slides

Venerable procedures for privacy protection and data sharing within academia, companies, and governments, and between sectors, have been proven to be completely inadequate (e.g., respondents in de-identified surveys can usually be re-identified). At the same time, unprecedented quantities of data that could help social scientists understand and ameliorate the challenges of human society are presently locked away inside companies, governments, and other organizations, in part because of worries about privacy violations. We address these problems with a general-...

Read more about Statistically Valid Inferences from Privacy Protected Data (University of Wisconsin)
Statistically Valid Inferences from Privacy Protected Data (Deloitte), at Deloitte, Thursday, July 14, 2022:
Slides

Venerable procedures used for privacy protection in sharing data within individual companies and governments, within academia, and between sectors have recently been proven massively inadequate (e.g., respondents in de-identified surveys can usually be re-identified). Furthermore, the benefits of getting our data sharing act together go far beyond the university, since unprecedented quantities of data that could help social scientists understand and ameliorate the challenges of human society are presently locked away inside companies, governments, and other organizations,...

Read more about Statistically Valid Inferences from Privacy Protected Data (Deloitte)
Statistically Valid Inferences from Privacy Protected Data (SICSS, University of Rochester), at SICSS, University of Rochester, Monday, May 9, 2022:
Slides

Venerable procedures used for privacy protection in sharing academic data have recently been proven massively inadequate (e.g., respondents in de-identified surveys can usually be re-identified). Furthermore, the benefits of getting our data sharing act together go far beyond the university, since unprecedented quantities of data that could help social scientists understand and ameliorate the challenges of human society are presently locked away inside companies, governments, and other organizations, in part because of worries about privacy violations. We address these...

Read more about Statistically Valid Inferences from Privacy Protected Data (SICSS, University of Rochester)
Public Policy for the Poor? A Randomized Evaluation of the Mexican Universal Health Insurance Program (Harvard School of Public Health), at Seminar on Health System Quality, Prof Margaret Kruk, Thursday, May 5, 2022:
Slides
An evaluation of the Mexican Seguro Popular program (designed to extend health insurance and regular and preventive medical care, pharmaceuticals, and health facilities to 50 million uninsured Mexicans), one of the world's largest health policy reforms of the last two decades. Our evaluation features a new design for field experiments that is more robust to the political interventions and implementation errors that have ruined many similar previous efforts; new statistical methods that produce more reliable and efficient results using fewer resources, assumptions, and data; and an... Read more about Public Policy for the Poor? A Randomized Evaluation of the Mexican Universal Health Insurance Program (Harvard School of Public Health)
Statistically Valid Inferences from Privacy Protected Data (Institute for Analytical Sociology, Norrköping, Sweden), at Institute for Analytical Sociology, Norrköping, Sweden, Wednesday, March 2, 2022:
Slides

Common procedures used for privacy protection in sharing academic data have now been proven massively inadequate (e.g., respondents in de-identified surveys can usually be re-identified). Furthermore, the benefits of getting our data sharing act together go far beyond the university, since unprecedented quantities of data that could help social scientists understand and ameliorate the challenges of human society are presently locked away inside companies, governments, and other organizations, in part because of worries about privacy violations. We address these ...

Read more about Statistically Valid Inferences from Privacy Protected Data (Institute for Analytical Sociology, Norrköping, Sweden)
Simplifying Matching Methods for Causal Inference (University of Wisconsin at Madison), at University of Wisconsin at Madison, Department of Population Health Sciences, Monday, February 7, 2022:
Slides
We show how to use matching methods for causal inference to ameliorate model dependence -- where small, indefensible changes in model specification have large impacts on our conclusions. We introduce methods that are simpler, more powerful, and easier to understand than existing approaches. We also show that propensity score matching, an enormously popular approach, accomplishes the opposite of its intended goal -- increasing imbalance, inefficiency, model dependence, and bias -- and should be replaced with other matching methods in applications.  See ... Read more about Simplifying Matching Methods for Causal Inference (University of Wisconsin at Madison)

Pages