*Dear all,
Please join us for the Applied Statistics Workshop (Gov 3009) this
Wednesday, September 26 from 12.00 - 1.30 pm in CGIS Knafel Room 354. Luke
Miratrix, Assistant Professor of Statistics in the Department of Statistics
at Harvard University, will give a presentation entitled "Random Weight
Estimators: Adjusting Randomized Trials Without Using Observed Outcomes". As
always, a light lunch will be provided.
Abstract:*
To increase the precision of a randomized trial, experimenters often adjust
> estimates of treatment effects using baseline covariates thought to predict
> the outcome of interest. In a previous paper, we proved that even under the
> Neyman-Rubin model, if the covariates and the method for adjustment are
> determined before randomization, this process can increase precision in a
> manner quite similar to a comparable blocked experiment. Typically,
> however, experimenters wish to adjust for the covariates that are most
> imbalanced between treatment and control, given the realized randomization.
> This leads to a much vexed variable selection problem that depends on the
> observed treatment assignment. To understand the issues behind this
> process, we examine a class of estimators we call "Random Weight
> Estimators" that adjust treatment effect estimates by weighting units with
> weights depending on a function on treatment assignment and covariates.
> While similar in spirit to blocking, these estimators can be applied "after
> the fact,'' i.e., after randomization has occurred, allowing them to
> naturally adapt to the observed treatment assignment. They can also adjust
> for many different covariates at once, including continuous ones. This
> class is quite general, and it includes traditional methods such as
> ordinary linear regression. Using our framework, we show, under the
> Neyman-Rubin model, how one can easily introduce potential bias using what
> would seem to be legitimate and simple approaches, especially in small and
> midsize experiments. Care must be taken with many forms of adjustment, even
> if an approach is selected without regard to any actual outcomes. We also
> extend this methodology to survey experiments, giving an appropriate and
> near-unbiased estimator for the treatment effect of a parent population.
> Throughout the talk, we illustrate this overall framework.
**
**
An up-to-date schedule for the workshop is available at
http://www.iq.harvard.edu/events/node/1208.
**
**
Best,
Konstantin
--
Konstantin Kashin
Ph.D. Candidate in Government
Harvard University
Mobile: 978-844-0538
E-mail: kkashin(a)fas.harvard.edu
Site: http://www.konstantinkashin.com/<http://people.fas.harvard.edu/%7Ekkashin/>
Dear all,
Please join us for the Applied Statistics Workshop (Gov 3009) this
Wednesday, September 19 from 12.00 - 1.30 pm in CGIS Knafel Room 354. Rich
Nielsen <http://people.fas.harvard.edu/~rnielsen/>, a Ph.D. candidate from
the Department of Government at Harvard University, will give a practice
job talk entitled "Jihadi Radicalization of Muslim Clerics". As always, a
light lunch will be provided.
Abstract:
**
> This paper explains why some Muslim clerics adopt the ideology of militant
> Jihad while others do not. I argue that clerics strategically adopt or
> reject Jihadi ideology because of career incentives generated by the
> structure of cleric educational networks. Well-connected clerics enjoy
> substantial success at pursuing comfortable careers within state-run
> religious institutions and they reject Jihadi ideology in exchange for
> continued material support from the state. Clerics with poor educational
> networks cannot rely on connections to advance through the state-run
> institutions, so many pursue careers outside of the system by appealing
> directly to lay audiences for support. These clerics are more likely to
> adopt Jihadi ideology because it helps them demonstrate to potential
> supporters that they have not been theologically coopted by political
> elites. I provide evidence of these dynamics by collecting and analyzing
> 29,430 fatwas, articles, and books written by 91 contemporary clerics.
> Using statistical natural language processing, I measure the extent to
> which each cleric adopts Jihadi ideology in their writing. I combine this
> with biographical and network information about each cleric to trace the
> process by which poorly-connected clerics become more likely to adopt
> Jihadi ideology.
An up-to-date schedule for the workshop is available at
http://www.iq.harvard.edu/events/node/1208.
Best,
Konstantin
--
Konstantin Kashin
Ph.D. Candidate in Government
Harvard University
Mobile: 978-844-0538
E-mail: kkashin(a)fas.harvard.edu
Site: http://www.konstantinkashin.com/<http://people.fas.harvard.edu/%7Ekkashin/>
Hi all,
Rich Nielsen indicated that he found a " white plug-to-usb device in the
back center plug of K354" (and that it's likely for an iphone). If you are
missing this, please email him directly at nielsen.rich(a)gmail.com.
-Konstantin
--
Konstantin Kashin
Ph.D. Candidate in Government
Harvard University
Mobile: 978-844-0538
E-mail: kkashin(a)fas.harvard.edu
Site: http://www.konstantinkashin.com/<http://people.fas.harvard.edu/%7Ekkashin/>
Dear all,
Please join us for the Applied Statistics Workshop (Gov 3009) this
Wednesday, September 12 from 12.00 - 1.30 pm in CGIS Knafel Room 354. Jamie
Robins <http://www.hsph.harvard.edu/faculty/james-robins/>, Professor of
Epidemiology at the Harvard School of Public Health, will give a
presentation entitled "A Simple Unification of the Potential Outcome and
Causal Graph Approaches to Causal Inference". As always, a light lunch will
be provided.
Abstract:
**
> Potential outcomes are extensively used within statistics, epidemiology,
> and political science for reasoning about causation. Directed acyclic
> graphs are another formalism used to represent causal systems. They are
> extensively used in computer science, bioinformatics, sociology and
> epidemiology. It is natural to wish to unify them.
We present a simple approach to this unification. The approach is based on
> the idea of splitting nodes to construct graphs whose nodes are potential
> outcomes. The resulting graph can be used to read off counterfactual
> independencies. These independencies are satisfied by all previously
> proposed graphical and nongraphical causal models. We review many examples
> to illustrate the power of this approach.
This is joint work with Thomas Richardson at the University of Washington.
An up-to-date schedule for the workshop is available at
http://www.iq.harvard.edu/events/node/1208.
Best,
Konstantin
--
Konstantin Kashin
Ph.D. Candidate in Government
Harvard University
Mobile: 978-844-0538
E-mail: kkashin(a)fas.harvard.edu
Site: http://www.konstantinkashin.com/<http://people.fas.harvard.edu/%7Ekkashin/>
Dear all,
Please join us for the first session of the Applied Statistics Workshop
(Gov 3009) this Wednesday, Sept. 5 from 12.00 - 1.30 pm in CGIS Knafel Room
354. Michael Grubb <http://www.mit.edu/~mgrubb/>, an Assistant Professor of
Applied Economics from the MIT Sloan School of Management will give a
presentation entitled "Cellular Service Demand: Biased Beliefs, Learning,
and Bill Shock <http://www.mit.edu/~mgrubb/GrubbOsborne.pdf>". As always, a
light lunch will be provided.
Abstract:
By April 2013, the FCC's recent bill-shock agreement with cellular carriers
> requires consumers be notified when exceeding usage allowances. Will the
> agreement help or hurt consumers? To answer this question, we estimate a
> model of consumer plan choice, usage, and learning using a panel of
> cellular bills. Our model predicts that the agreement will lower average
> consumer welfare by $2 per year because firms will respond by raising
> monthly fees. Our approach is based on novel evidence that consumers are
> inattentive to past usage (meaning that bill-shock alerts are informative)
> and advances structural modeling of demand in situations where multi-part
> tariffs induce marginal-price uncertainty. Additionally, our model
> estimates show that an average consumer underestimates both the mean and
> variance of future calling. These biases cost consumers $42 per year at
> existing prices. Moreover, absent bias, the bill-shock agreement would have
> little to no effect.
Note that this work is joint with Matthew Osborne at the Bureau of Economic
Analysis.
An up-to-date schedule for the workshop is available at
http://www.iq.harvard.edu/events/node/1208.
Best,
Konstantin
--
Konstantin Kashin
Ph.D. Candidate in Government
Harvard University
Mobile: 978-844-0538
E-mail: kkashin(a)fas.harvard.edu
Site: http://www.konstantinkashin.com/<http://people.fas.harvard.edu/%7Ekkashin/>