Hi everyone!
At the applied statistics workshop this week (Wednesday, October 29th) we
have *Maggie McConnell, *a professor at the Harvard School of Public
Health. She will present work entitled *To Charge or Not to Charge:
Evidence from a Health Products Experiment in Uganda. *Please find the
abstract below as well as on the website (here
<http://projects.iq.harvard.edu/applied.stats.workshop-gov3009/presentations…>
).
As usual, we will meet in CGIS Knafel Room 354 from 12-1:30 pm. Lunch will
be provided. See you there!
-- Dana Higgins
*Title: *To Charge or Not to Charge: Evidence from a Health Products
Experiment in Uganda
*Abstract:* In a field experiment in Uganda, we find that demand after a
free distribution of three health products is lower than after a sale
distribution. This contrasts with earlier work on insecticide-treated
bednets, highlighting the importance of considering product characteristics
in determining pricing policy. We put forward a model to illustrate the
potential tension between two important factors, learning and anchoring,
and then test this model with three products selected specifically for
their variation in the scope for learning. We find the rank order of shifts
in demand matches with the theoretical prediction, although differences are
not statistically significant across all products.
Sorry for the SPAM but I think many of you might find today's talk in the PPE Seminar very interesting.
http://www.sas.upenn.edu/~gcamilo/cgj/Research_files/network%20state%20capa…
The paper has an interesting structural estimation that could be applied to other questions.
Best,
Horacio
-------------------------------------
Horacio Larreguy Arbesu
Assistant Professor of Government | Harvard University
1737 Cambridge St, CGIS Knafel Building RM 408, Cambridge MA 02138
+1 617 496 1497
________________________________
From: junior-bounces(a)lists.gov.harvard.edu [junior-bounces(a)lists.gov.harvard.edu] on behalf of Stone, Gabrielle [gstone(a)iq.harvard.edu]
Sent: Thursday, October 23, 2014 7:23 AM
To: ppe_list(a)lists.iq.harvard.edu
Cc: Camilo Garcia-Jimeno (gcamilo(a)sas.upenn.edu)
Subject: [junior] [ppe_list] Postive Political Economy Series Today
Please join us on Thursday, October 23 at 4:30pm (Room K354 of CGIS Knafel, 1737 Cambridge St.) for the next Positive Political Economy Series of the Fall 2014 semester.
University of Pennsylvania Professor Camilo Garcia-Jimeno, will present, "State Capacity and Economic Development: A Network Approach."
The Faculty of Arts and Sciences and The Institute for Quantitative Social Science at Harvard University are sponsoring a seminar on formal and quantitative political research. The Program on Positive Political Economy (PPE) supports research-related activities that integrate the study of economics and politics, whether by studying economic behavior in the political process or political behavior in the marketplace. In general, positive political economy is concerned with showing how observed differences among institutions affect political and economic outcomes in various social, economic, and political systems and how the institutions themselves change and develop in response to individual and collective beliefs, preferences, and strategies. All interested faculty and students are invited to attend.
For more information, please go to: http://www.iq.harvard.edu/event/harvard-seminar-positive-political-economy-….
Gabrielle Stone
Events Coordinator
Institute for Quantitative Social Science (IQSS)
1737 Cambridge Street
Room K315
Cambridge, MA 02138
617-495-9489
gstone(a)iq.harvard.edu<mailto:gstone@iq.harvard.edu>
OFFICE HOURS: M, T, Th 8:30am-4:30pm, out on W, Fr
Hi everyone,
Sorry for the second email, but there was a mistake made in the
announcement.
Our speaker this Wednesday will be Victor Chernozhukov, who is a Professor
of Economics at MIT. Victor's talk is Gaussian Approximations, Bootstrap,
and Z-estimators when p>>n.
Abstract is correct. I apologize for the mixup!
-- Dana Higgins
Hi everyone!
Our speaker this Wednesday (10/22) at Applied Stats will be *Victor
Chernozhukov, *who will be practicing his job talk. Brandon will be giving
a talk entitled *Gaussian Approximations, Bootstrap, and Z-estimators when
p>>n. *The abstract for the talk is included below. As per usual, we will
meet in CGIS K354 at noon and lunch will be served.
I look forward to seeing you all there! Thank you!
-- Dana Higgins
Title: Gaussian Approximations, Bootstrap, and Z-estimators when p >> n.
Abstract: We show that central limit theorems hold for high-dimensional
normalized means hitting high-dimensional rectangles. These results apply
even when p>> n. These theorems provide Gaussian distributional
approximations that are not pivotal, but they can be consistently estimated
via Gaussian multiplier methods and the empirical bootstrap. These results
are useful for building confidence bands and for multiple testing via the
step-down methods. Moreover, these results hold for approximately linear
estimators. As an application we show that these central limit theorems
apply to normalized Z-estimators of p> n target parameter in a class of
problems, with estimating equations for each target parameter
orthogonalized with respect to the nuisance functions being estimated via
sparse methods. (This talk is based primarily on the joint work with Denis
Chetverikov and Kengo Kato.)
Hi everyone!
This week we have *Teppei Yamamoto, *a Professor of Political Science at
MIT. He will be giving a talk entitled *Design, Identification, and
Sensitivity Analysis for Patient Preference Trials. *The abstract for the
project is included below and available on the website (here
<http://projects.iq.harvard.edu/applied.stats.workshop-gov3009/home>). As
usual, we will meet in CGIS K354 at noon and lunch will be served.
I look forward to seeing you all there! Thanks!
-- Dana Higgins
Title:
Design, Identification, and Sensitivity Analysis for Patient Preference
Trials
Authors:
Dean Knox, Teppei Yamamoto, Matthew A. Baum, and Adam Berinsky
Abstract:
Social and medical scientists are often concerned that the external
validity of experimental results may be compromised because of
heterogeneous treatment effects. If a treatment has different effects on
those who would choose to take it and those who would not, the average
treatment effect estimated in a standard randomized controlled trial (RCT)
may give a misleading picture of its overall impact outside of the study
sample. Patient preference trials (PPTs), where participants' preferences
over treatment options are incorporated in the study design, provide a
possible solution. In this paper, we provide for the first time a
systematic analysis of PPTs based on the potential outcomes framework of
causal inference. We propose a general design for PPTs with multi-valued
treatments, where participants state their preferred treatments and are
then randomized into either a standard RCT or a self-selection condition.
We derive nonparametric bounds on the average causal effects among each
choice-based subpopulation of participants under the proposed design.
Finally, we propose a sensitivity analysis for the violation of the key
ignorability assumption sufficient for identifying the target causal
quantity. The proposed design and methodology are illustrated with an
original study of partisan news media and its behavioral impact.
Hi everyone!
This week we have *Alberto Abadie, *a Professor of Public Policy at the
Harvard Kennedy School. He will be giving a talk entitled *Endogenous
Stratification in Randomized Experiments*. The abstract for the project is
included below and available on the website (here
<http://projects.iq.harvard.edu/applied.stats.workshop-gov3009/presentations…>).
As usual, we will meet in CGIS K354 at noon and lunch will be served.
I look forward to seeing you all there! Thanks!
-- Dana Higgins
*Authors:* Alberto Abadie, Matthew M. Chingos, and Martin R. West
*Abstract:* Researchers and policy makers are often interested in
estimating how treatments or policy interventions affect the outcomes of
those most in need of help. This concern has motivated the increasingly
common practice of disaggregating experimental data by groups constructed
on the basis of an index of baseline characteristics that predicts the
values that individual outcomes would take on in the absence of the
treatment. This article shows that substantial biases may arise in practice
if the index is estimated, as is often the case, by regressing the outcome
variable on baseline characteristics for the full sample of experimental
controls. We analyze the behavior of leave-one-out and repeated split
sample estimators and show they behave well in realistic scenarios,
correcting the large bias problem of the full sample estimator. We use data
from the National JTPA Study and the Tennessee STAR experiment to
demonstrate the performance of alternative estimators and the magnitude of
their biases.
