Social Statistics Seminar Series
Tue, 12/10/2013, 12:00 PM—1:30 PM
4357 Bunche Hall
Brenton Kenkel
Politics
Princeton University
A General Solution to Nonignorable Missing Outcomes in Binary Choice Data
I provide a general method to obtain logistic regression coefficients when there is nonignorable
missingness in the response variable. The method makes no assumptions about the missingness
mechanism and requires no application-specifi c programming. The key is to treat the model
parameters as partially identifi ed and estimate bounds on the set of regression coefficients that
are consistent with the data, as opposed to the traditional approach of obtaining a single point
estimate. Drawing from the partial identi fication literature in econometrics, I develop three
closely related methods for estimating bounds on the parameters of a logistic regression model
when the missingness mechanism is wholly unknown to the analyst. These methods are computationally intensive but straightforward to implement. The fi rst entails sampling from the
set of possible "completed" outcome vectors, while the latter two are based on criteria used for
similar models in econometrics. I run a simulation to analyze the performance of each method
across varying sample sizes and find that one of the criterion-based estimators, the minimum
distance method, is generally most accurate at recovering the set of observationally equivalent
parameters. I apply the method to Lyall's (2010) data on insurgency outcomes and fi nd that
the original results are robust to assumptions about the treatment of "draws."
Statistics/Social Sciences Tenure Track Faculty Position
Open-ranked tenured and tenure track position, joint in the Department of Statistics and a department or departments in the Division of Social Sciences at the University of California, Los Angeles. We seek a faculty member contributing at the cutting edge to the development of statistical methodology relevant to the social sciences. We welcome candidates whose experience in teaching, research or community service has prepared them to contribute to our commitment to diversity and excellence. Duties include new course development, teaching, and methodological and collaborative research. Ph.D. required by date of anticipated appointment of July 1, 2014. Salary will be commensurate with education and experience. Please refer to Tracking Job #1015-1314-01 on all correspondence.
Reviews for the position begin October 1, 2013, and will continue until the position is filled. Interested applicants should apply at UCLA Academic Recruit at:
https://recruit.apo.ucla.edu/apply/JPF00072
Applicants should submit:
1) CV
2) Cover letter describing how their qualifications and interests would fit with the position description
3) Statement of Research
4) Statement of Teaching
5) Three letters of recommendation
For more information on the departments in the Division of Social Sciences see www.sscnet.ucla.edu.
All inquiries may be sent to Professor Mark Handcock at css@stat.ucla.edu.
UCLA is an Affirmative Action/Equal Opportunity Employer with a strong institutional commitment to excellence through diversity. Women and Minorities are especially encouraged to apply.
Inferring "Dark Matter" from Videos
Our VCLA (Vision, Cognition, Learning and Art) research group built two low cost hex rotor helicopters to take aerial videos on Sep 11, 2013 in Malibu State Park. These videos captured the motion of 15 "actors" from our group playing different roles in various activities. Designing and carrying out the activities required creativity and was a lot of fun for the participants.
In the aerial videos it is difficult to identify some objects due to view angle, distance or occlusions (without considering other cues). These unidentified objects are called “dark matter”. The purpose of this exercise was to collect data to test a novel model based on the cue of human activities in the scene. This model, proposed by Dan Xie, Sinisa Todorovic and Song-Chun Zhu, will appear in Proceedings of International Conference on Computer Vision (ICCV) 2013.
UCLA Statistics Faculty Search
Open-ranked faculty position (tenured or tenure track) in the Department of Statistics at the University of California, Los Angeles. We seek applicants contributing to the cutting edge development of theoretical statistics, learning theory and big data. We welcome applicants whose experience in research, teaching and service has prepared them to contribute to UCLA’s commitment to excellence and diversity. Applicants with a desire to mentor women and/or underrepresented minorities in STEM fields are encouraged to apply. Ph.D. is required by date of anticipated appointment of July 1, 2014. Salary will be commensurate with education and experience. Please refer to Tracking Job #1015-1314-02 on all correspondence.
Reviews for the position will commence on Nov 15, 2013 and will continue until a successful candidate is identified. Interested applicants should apply at UCLA Academic Recruit at:
https://recruit.apo.ucla.edu/apply/JPF00108
UCLA is an Affirmative Action/Equal Opportunity Employer with a strong institutional commitment to excellence through diversity. Women and Minorities are especially encouraged to apply.
Applicants should submit:
1) CV
2) Cover letter describing how their qualifications and interests would fit with the position description
3) Statement of research
4) Three representative publications
5) Statement of teaching
6) Three letters of recommendation
All inquiries may be sent to Chie Ryu at chie.ryu@stat.ucla.edu.
Announcing the R User Conference 2014 to be held from June 30 to July 3 2014 at UCLA
The annual useR! international R User conference is the main meeting of the R user and developer community. It is being held at UCLA from June 30 to July 3, 2014. Please goto http://user2014.stat.ucla.edu for more details.