Daniel Lee June 23, 2015 10:00 AM-12:00 PM 4240 Public Affairs Building Stan is an open-source, Bayesian inference tool with interfaces in R, Python, Matlab, Julia, Stata, and the command line. Users write statistical models in a high-level statistical language. The default Bayesian inference algorithm is the no-U-turn sampler (NUTS), an auto-tuned version of Hamiltonian […]
"Inferring and understanding travel and migration movements at a global scale"
Abstract: Despite extensive work on the dynamics and outcomes of large-scale migrations, timely and accurate estimates of population movements do not exist. While censuses, surveys, and observational data have been used to measure migration, estimates based on these data sources are constrained in their inability to detect unfolding migrations, and lack temporal and demographic detail. In this study, we present a novel approach for generating estimates of migration that can measure movements of particular demographic groups across country lines.
Specifically, we model migration as a function of long-term moves across countries using aggregated Facebook data. We demonstrate that this methodological approach can be used to produce accurate measures of past and ongoing migrations - both short-term patterns and long-term changes in residence. Several case studies confirm the validity of our approach, and highlight the tremendous potential of information obtained from online platforms to enable novel research on human migration events.
Presented By: Mark S. Handcock (Professor, Statistics) Jeffrey B. Lewis (Professor, Political Science) Marc A. Suchard (Professor, Biomathematics, Biostatistics and Human Genetics) Reproducibility is one of the main principles of the scientific method. This panel of scholars will discuss issues in the importance of replication of statistical results. Increasing attention is being paid to […]
"Electronic Homestyle: Tweeting Ideology"
Abstract: Ideal points are central to the study of political partisanship and an essential component to our understanding of legislative and electoral behavior. We employ automated text analysis on tweets from Members of Congress to estimate their ideal points using Naive Bayes classification and Support Vector Machine classification. We extend these tools to estimate the proportion of partisan speech used in each legislator's tweets. We demonstrate an association between these measurements, existing ideal point measurements, and district ideology.
"Quantifying the dynamics of multimodal communication with multimodal data."
*Presented by the Center for Social Statistics
Abstract: Human communication is built upon an array of signals, from body movement to word selection. The sciences of language and communication tend to study these signals individually. However, natural human communication uses all these signals together simultaneously, and in complex social systems of various sizes. It is an open puzzle to uncover how this multimodal communication is structured in time and organized at different scales. Such a puzzle includes analysis of two-person interactions. It also involves an understanding of much larger systems, such as communication over social media at an unprecedentedly massive scale.
Collaborators and I have explored communication across both of these scales, and I will describe examples in the domain of conflict. For example, we've studied conflict communication in two-person interactions using video analysis of body and voice dynamics. At the broader scale, we have also used large-scale social media behavior (Twitter) during a massively shared experience of conflict, the 2012 Presidential Debates. These projects reveal the importance of dynamics. In two-person conflict, for example, signal dynamics (e.g., body, voice) during interaction can reveal the quality of that interaction. In addition, collective behavior on Twitter can be predicted even by simple linear models using debate dynamics between Obama and Romney (e.g., one interrupting the other).
The collection, quantification, and modeling of multitemporal and multivariate datasets hold much promise for new kinds of interdisciplinary collaborations. I will end by discussing how they may guide new theoretical directions for pursuing the organization and temporal structure of multimodality in communication.
"Innovative Sampling Approaches for Hard to Reach Populations: Design of a National Probability Study of Lesbians, Gay Men, Bisexuals, and Transgender Peoples and Network Sampling of Hard to Reach Populations"
Speakers:
Ilan H. Meyer, Williams Distinguished Senior Scholar for Public Policy at the Williams Institute
Mark S. Handcock, Professor of Statistics at UCLA and Director of the Center for Social Statistics
Description:
Come for the exciting seminar then stay for the free lunch and discussion. A seminar led by Ilan H. Meyer followed immediately by a Brown Bag Lunch led by Mark S. Handcock.
Dr. Meyer is Principal Investigator of the Generations and TransPop Surveys. Generations is a survey of a nationally representative sample of 3 generations of lesbians, gay men, and bisexuals. TransPop is the first national probability sample survey of transgender individuals in the United States. Both studies attempt to obtain large nationally representative samples of hard to reach populations. Dr. Meyer will review sampling issues with LGBT populations and speak on the importance of measuring population health of LGBTs and the underlying aspects in designing a national probability survey.
From a contrasting perspective, the field of Survey Methodology is facing many challenges. The general trend of declining response rates is making it harder for survey researchers to reach their intended population of interest using classical survey sampling methods.
In the followup Brown Bag Lunch, led by Mark S. Handcock, participants will discuss statistical challenges and approaches to sampling hard to reach populations. Transgenders, for example, are a rare and stigmatized population. If the transgender community exhibits networked social behavior, then network sampling methods may be useful approaches that compliment classical survey methods.
Participants are encouraged to speak on ideas of statistical methods for surveys.
