Skip to content

Data Science Seminar: Andrew Gelman, “The Statistical Crisis in Science”

Top journals routinely publish ridiculous, scientifically implausible claims, justified based on “p < 0.05.” And this in turn calls into question all sorts of more plausible, but not necessarily true, claims, that are supported by this same sort of evidence. To put it another way: we can all laugh at studies of ESP, or ovulation and voting, but what about MRI studies of political attitudes, or stereotype threat, or, for that matter, the latest potential cancer cure? If we can’t trust p-values, does experimental science involving human variation just have to start over? And what do we do in fields such as political science and economics, where preregistered replication can be difficult or impossible? Can Bayesian inference supply a solution? Maybe. These are not easy problems, but they’re important problems.

Andrew Gelman is a professor of statistics and political science and director of the Applied Statistics Center at Columbia University. He has received the Outstanding Statistical Application Award from the American Statistical Association, the award for best article published in the American Political Science Review, and the Council of Presidents of Statistical Societies Award for outstanding contributions by a person under the age of 40. His books include Bayesian Data Analysis (with John Carlin, Hal Stern, David Dunson, Aki Vehtari, and Don Rubin), Teaching Statistics: A Bag of Tricks (with Deb Nolan), Data Analysis Using Regression and Multilevel/Hierarchical Models (with Jennifer Hill), Red State, Blue State, Rich State, Poor State: Why Americans Vote the Way They Do (with David Park, Boris Shor, and Jeronimo Cortina), and A Quantitative Tour of the Social Sciences (co-edited with Jeronimo Cortina).

Emilio Zagheni Leverages Facebook as a Tool for Demographic Research

In an article published last week in Population and Development Review, CSDE Training Director Emilio Zagheni and co-authors present Facebook’s advertising platform as a novel strategy for estimating migration data. Facebook provides a free and accessible means for researchers and advertisers alike to find relevant socio-demographic information about users, which they can then sort by geographic location. Considering the dearth of data sources available for assessing progress towards goals such as sustainable development, this approach—which takes advantage of new data sources—is an important development for demographic research. You can read more about their innovative approach below.

Eleanor Brindle Develops Method for Assessing Micronutrient Deficiencies in Low- and Middle-Income Countries

In a recent article, CSDE Biodemography Lab Manager Eleanor Brindle and colleagues present a technique to address the issue of limited information on the status micronutrient deficiencies in the populations of low- and middle-income countries. The micronutrients of concern to the authors are vitamin A, iron, and iodine, deficiencies of which are public health concerns in a number of low- and middle-income nations. Lead author Brindle and her co-authors developed a 7-plex immunoassay that enables the concurrent measurement of seven biomarkers used to determine the status of these micronutrients in populations of concern. According to the authors, this technique could provide an affordable means of surveying populations for nutrient deficiencies, as well as for falciparum malarial parasitemia inflammation and infection. The article is a product of work Brindle has been doing with the international nonprofit organization PATH over past few years, and is accessible below.

Janelle Taylor Receives Funding from National Institute on Aging for Study on Health Outcomes for Dementia Patients

Janelle S. Taylor, affiliate and Professor in the Department of Anthropology, has received funding from the National Institute on Aging (NIA) for a new study titled “Health Outcomes for Patients with Dementia without Family Caregivers.”  This two-year R21 project includes both quantitative and qualitative components, and will be a collaboration between Taylor as PI and colleagues from Kaiser Permanente Washington Health Research Institute; the VA Puget Sound Health Care System; UW Medicine, Social Work, Anthropology; and CSDE. Collaborators include Marlaine F. Gray, Eric B. Larson, Paul K. Crane, Elizabeth K. Vig, Stephanie G.B. Wheeler, Ann M. O’Hare, Clara W. Berridge, and Bettina Shell-Duncan, along with CSDE’s Cori Mar. The research will analyze data from the Adult Changes in Thought (ACT) study–an ongoing longitudinal cohort study of incident dementia–in combination with health records data, to document whether older adults with dementia who lack family support suffer worse health outcomes, and/or different patterns of interaction with the health care system, than do those with strong family support.

This is Taylor’s first time submitting an NIH grant as PI, and it was funded on the first submission–a happy outcome for which Taylor gives much credit to the excellent support that she received from CSDE.  She relied upon CSDE staff not only to apply for the grant (John Kemner), but also for guidance on NIH policies and procedures (John Kemner, Sara Curran, Bettina Shell-Duncan), comments and feedback on application drafts (Sara Curran), and advice on study design and preliminary assessments of data (for example, when requesting preliminary data and preparing a power analysis; Cori Mar).  She hopes that her positive experience and outcome, as a first-time NIH applicant, may offer encouragement to other CSDE affiliates considering embarking upon what can be a rather daunting application process.

Data Science Incubation Program

The UW eScience Institute is pleased to announce
the Data Science Incubation Program – Winter 2018

The goal of the Data Science Incubator is to enable new science by bringing together data scientists and domain scientists to work on focused, intensive, collaborative projects.  Our team of data scientists provides expertise in state-of-the-art technology and methods in statistics and machine learning, data manipulation and analytics at all scales, cloud and cluster computing, software design and engineering, visualization, and other topics. We invite short proposals (1-2 pages) for one-quarter data-intensive research projects focusing on extracting insight from large, noisy, or heterogeneous datasets.

The program is open to any faculty, postdoc, staff, or student whose research can be significantly advanced by intensive collaboration with a data science expert. To apply, we require a short project proposal describing the science goals, the relevant datasets, and the expected technical challenges.  The ideal proposal will clearly identify both the datasets involved and the questions to be answered, and will explain how the technical component of the project is critical to delivering exciting new findings.

Each project must include a project lead who is willing to physically co-locate with the incubator staff. We find that collaboration in a shared space is important for deeper technical engagement and provides opportunities for “cross-pollination” among multiple concurrent projects. The Incubator operates on Tuesdays and Thursdays out of the  WRF Data Science Studio  (6th floor of the Physics/Astronomy Tower). The project lead should plan to work in the Studio for several hours on these days.

Incubator projects are not “for-hire” software jobs — the project lead will work in collaboration with the data scientists and the broader eScience community. Each project lead will “own” their project (and its results) and be responsible for its successful completion, with the eScience team providing guidance on methods, technologies, and best practices as well as general software engineering.

For more information including how to apply, see our Incubator Program page.

Important Dates for the Winter 2018 Incubator:

October 26th: Information meeting. Location: WRF Data Science Studio. Time: 1pm.

November 10th: Applications due.

December 4th: Notification or proposal selections.

January 4th: Kickoff meeting. Location: WRF Data Science Studio.

 

If you have any questions about submitting a project for an Incubator, please consult one of our Data Scientists during their Office Hours for guidance: http://escience.washington.edu/office-hours

 

Moving to Despair? Migration and Well-Being in Pakistan

Valerie Mueller, School of Politics and Global Studies, Arizona State University

Internal migration has the potential to substantially increase income, especially for the poor in developing countries, and yet migration rates remain low. We explore the role of psychic costs by evaluating the impacts of internal migration on a suite of well-being indicators using a unique, 22-year longitudinal study in rural Pakistan. We account for selection into migration using covariate matching. Migrants have roughly 35 to 40 percent higher consumption per adult equivalent, yet are 12 to 14 percentage points less likely to report feeling either happy or calm. Our results suggest that deteriorating physical health coupled with feelings of relative deprivation underlie the disparity between economic and mental well-being. Thus, despite substantial monetary gains from migration, people may be happier and less mentally distressed remaining at home. If traditional market mechanisms cannot reduce psychic costs, it may be more constructive to address regional inequality by shifting production – rather than workers – across space.

Lecturer in Population Studies/Demography

The School of Geography and Sustainable Development is seeking to appoint a Lecturer with expertise in Population Studies/Demography. We are looking for candidates with excellent research profiles in any area of population research. Candidates should have strong skills in demographic and quantitative data analysis and be capable of contributing to the Human Geography undergraduate teaching programme and to a planned taught postgraduate programme in Demography/Population Analysis.

Candidates should hold a PhD in demography, population geography or a cognate social science discipline (including statistics and applied mathematics). Excellent teaching skills and an interest in promoting knowledge exchange are essential. Familiarity with grant seeking processes in relation to research councils and other national and international sources will be an advantage. This is an excellent opportunity for an enthusiastic individual to work in and collaborate with an international team of population researchers on highly policy-relevant topics, exploiting population data and applying sound methodology.

Informal enquiries can be directed to: Professor Hill Kulu, Hill.Kulu@standrews.ac.uk, Professor Elspeth Graham, efg@st-andrews.ac.uk, or Professor William Austin, goshos@standrews.ac.uk.

Assistant Professor in Human Geography

This is an exciting opportunity to appoint outstanding individuals to three Assistant Professorships in Human Geography whose work will complement and extend the current work in any one (or more) of our research clusters (Economy and Culture; Geographies of Life; Politics – State – Space; Urban Worlds). We are looking for first-rate scholars in their field and we encourage work that develops a range of theoretical, conceptual and/or methodological approaches. Work with relevance to the topics of the Global Challenges identified by RCUK would be an advantage and we anticipate that one post will offer expertise in Geographic Information Systems.

Successful candidates will have a PhD in a relevant field, and demonstrate a track record of, or potential for, internationally excellent and world class agenda setting publications and securing external funding. They will show a commitment to delivering excellent teaching at undergraduate and postgraduate levels, the highest quality supervision to research postgraduates, and will be expected to contribute to undergraduate and masters-level teaching. In addition to their research, the successful candidates will perform departmental administrative tasks commensurate with their grade and contribute to our wider human geography teaching programme.

The closing date for applications is October 27, 2017. Successful applicants will, ideally, be in post by September 1, 2018. Please direct informal enquiries to Professor Cheryl McEwan (cheryl.mcewan@durham.ac.uk).

Assistant Professor of Migration Studies

New College of Florida invites applications for a tenure-track position as Assistant Professor focusing on Migration Studies. This position is open with regard to field of PhD, and the appointment will be in that field with the opportunity of a joint appointment in another field or interdisciplinary program.

Candidates should have a PhD in a relevant discipline, teaching experience, a career interest in liberal arts education, and a vibrant research agenda. The person filling this line will be expected to teach courses on migration and build and coordinate interdisciplinary initiatives related to this subject, as well as contribute to the curriculum in the field of appointment. Experience using and teaching GIS and a willingness to guide students in developing GIS research skills is highly valued. The candidate must have completed the PhD by the start of employment in August 2018.

Candidates should apply on-line and upload the following: a cover letter; c.v.; statements regarding teaching philosophy, the candidate’s contributions or approaches to supporting an inclusive culture and learning environment, and research interests; syllabi; teaching evaluations; one writing sample (article or chapter length); and unofficial graduate transcript(s). In addition, please ask at least three references to send signed letters on letterhead electronically to MigrationSearch@ncf.edu or in hard copy to:  Migration Search, Division of Social Sciences, New College of Florida, 5800 Bay Shore Rd, Sarasota, FL 34243.

Postdoctoral Associate – Cornell Population Center

The Cornell Population Center (CPC) invites applicants for the Frank H.T. Rhodes Postdoctoral Fellowships. The start date for the position will be August 16, 2018 and it will be funded for 2 years, subject to a satisfactory first year evaluation.  Selection will be based on scholarly potential, ability to work in multi-disciplinary settings, and the support of a CPC faculty affiliate at Cornell who will serve as mentor, working closely with the Postdoctoral Associate. Preference will be given to fellows with research interests in areas broadly related to the CPC’s four main foci: families & children; health behaviors & disparities; poverty & inequality; and immigration & diversity. Especially encouraged are applications from candidates whose research has significance for those countries on which the fellowship’s funder focuses – the United States, the Republic of Ireland, Northern Ireland, Vietnam, South Africa, and Bermuda.

Postdoctoral Associates will have access to university resources and receive an annual salary of $63,000 plus benefits and a research/travel account.  Applicants must have completed a PhD in Demography, Economics, Sociology, or another related social science discipline by August 16, 2018. Screening of applications begins December 8, 2017, and will continue until the position is filled.

Applications must include: (a) letter of application, (b) curriculum vita, (c) a statement proposing both an individual research project and how the candidate will engage with a CPC faculty affiliate’s on-going research, (d) examples of written work, (e) a letter from a CPC faculty affiliate agreeing to mentor the candidate, and (f) three letters of recommendation. These materials must be submitted online via Academic Jobs Online, https://academicjobsonline.org/ajo/jobs/10058.  For questions, please contact Meg Cole (mc129@cornell.edu).