University of Maryland

How Do We Know That Our Statistical Methods Should Work? Benchmarks, Plasmodes, and Statistical Mediation Analysis

MONDAY SYMPOSIUM IN MEASUREMENT AND STATISTICS (MSMS)  UNIVERSITY OF MARYLAND together with  OHIO STATE UNIVERSITY, UNIVERSITY OF NORTH CAROLINA AT CHAPEL HILL* UNIVERSITY OF NOTRE DAME (* Organizer of this talk)   This presentation describes a benchmark method to validate statistical methods from the analysis of data on a known or established empirical effect. There are aspects to benchmark validation […]

Who Should Stop Unethical A.I.?

SoDa Associate and UMD Faculty Katie Shilton quoted in the New Yorker In computer science, the main outlets for peer-reviewed research are not journals but conferences, where accepted papers are presented in the form of talks or posters. In June, 2019, at a large artificial-intelligence conference in Long Beach, California, called Computer Vision and Pattern […]

Testing Remote Recruitment in a Television Measurement Panel (Feb 10, 2021)

Mixed-mode contact and recruitment approaches have become increasingly critical in reaching people to participate in surveys and research panels and improving their likelihood of response and often, the representation of the sample, bringing groups more traditionally represented. Nielsen has designed a phased sequential multi-mode methodology to recruit homes through a combination of web, phone, and in-person/proximity methods for its core TV panels that have relied exclusively on in-person recruitment. This work has been accelerated due to the COVID-19 pandemic. Nielsen will present the results of a “push to web” recruitment test conducted in five markets from October 2020 to January 2021.

The Nature and Impact of Hidden Data Errors on Information Risk and Data Science

Information Risk is an important field that encompasses multiple existing disciplines and overcomes the boundaries among them that has impeded knowledge sharing and effective management of integrated business projects. These projects often span multiple organizational groups and use different structured frameworks of information, data, computing, and security management. The practical realities intrinsic to performing this work leads to gaps in complying with regulations, ensuring secure operations, satisfying auditors, and even meeting program objectives for data analysis and business uses. This talk describes how deeply embedded data disparities that remain hidden to typical data methods lead to high error rates in project results. Lessons learned from assessing and correcting these situations is presented with examples of the problems and methods to detect and fix them.

Distance in Spatial Analysis: Challenges Related to Spatial Data Aggregation, Scale, and Computation

Distance is one of the most critical concepts in geography and has been widely used to quantify spatial separation between geographical entities. While measuring the distance between two points is straightforward, assessing the spatial separation between non-point objects can be challenging. This study investigates distance measurement between a location (point) and an area (polygon).

Bias Propensity to Inform Responsive and Adaptive Survey Design

Responsive and adaptive survey designs can be used to reduce the risk of nonresponse bias through data collection. In responsive design, different protocols that appeal to prior nonrespondents can be introduced in phases. In adaptive survey design, particular nonrespondents can be targets in these subsequent phases, based on predefined criteria. In the case of nonresponse bias, the criteria can be propensity models. Key, however, it the specification of these models.

Large-Scale Infrastructure for Social Data Science

Gathering, managing, and using social data to address critical questions requires the development of large-scale data infrastructures. In this Social Data Science Center (SoDa) panel, the distinguished speakers will share their experiences with the challenges and opportunities presented by efforts to create social data science infrastructure at the state, national, and international level.

MIS Quarterly Special Issue Showcase

Next-Generation Information Systems Theory January 27, January 29, and February 4th, 2021 Accelerating change, increasing complexity, and the unprecedented availability of data and algorithms for pattern identification have led some to argue for a reduced emphasis on theory in IS research.  However, it is our contention that theorizing is now more critical than ever. Rather […]

Data on Economic Anxiety Offer New Opportunities for Insights on the Global Effects of the COVID-19 Pandemic

Frauke Kreuter, Esther Kim, Sarah LaRocca, Katherine Morris, Christoph Kern, Andres Garcia December 21, 2020 Context The Global COVID-19 Trends and Impact Survey, which launched in April, 2020, is currently the largest ongoing public health data collection effort related to COVID-19. The survey is led by the University of Maryland and Carnegie Mellon University in […]

Measuring Emotion, Conflict and Disagreement – Watch Event Video

Social Data Science Center presents: Measuring Emotion, Conflict and Disagreement A Panel Discussion with Q&ATuesday, November 10, 2020 Panelists Annotation of Emotions and DisagreementPresented By: Dr. Susannah B. F. PaletzResearch Professor, College of Information Studies (iSchool), University of Maryland.Affiliate, Applied Research Lab for Intelligence and Security (ARLIS).Associate, Social Data Science Center (SoDa). When Does Disagreement […]