Dear Applied Statistics Workshop Community,
Our next meeting of the semester will be on October 5 (12:00 EST). Nima
Hejazi will present "Evaluating treatment efficacy in vaccine clinical
trials with two-phase designs using stochastic-interventional causal
effects."
<Where>
In-person: CGIS K354
Bagged lunches are available for pick-up at 11:40 (CGIS K354).
Zoom:
https://harvard.zoom.us/j/99181972207?pwd=Ykd3ZzVZRnZCSDZqNVpCSURCNnVvQT09
<Abstract>
In clinical trials randomizing participants to active vs. control
conditions and following study units until the occurrence of a primary
clinical endpoint, evaluating the efficacy of a quantitative exposure
(e.g., drug dosage, drug-induced biomarker activity) is often challenging,
as statistical innovations in causal inference have historically focused on
estimands compatible only with binary or categorical exposures.
Stochastic-interventional effects, which measure the causal effect
attributable to perturbing the exposure's natural (i.e., observed) value,
provide an interpretable solution. Unfortunately, their use in vaccine
efficacy trials requires extra care, for such trials measure immunologic
biomarkers – useful for understanding the mechanisms by which vaccines
confer protection or as surrogate endpoints in future trials – via
outcome-dependent two-phase sampling (e.g., case-cohort) designs. These
biased, outcome-dependent sampling designs have earned their popularity:
they circumvent the administrative burden of collecting potentially
expensive biomarker measurements on all study units without limiting
opportunities to detect important biomarkers that may be mechanistically
informative of the disease or infection process. We outline a
semiparametric biased sampling correction that allows for asymptotically
efficient inference on a causal vaccine efficacy measure defined by
contrasting assignments of study units to active vs. control while
simultaneously hypothetically shifting biomarker expression in the active
condition, yielding a causal dose-response analysis informative of
next-generation vaccine efficacy and useful for transporting efficacy from
a source pathogen strain (e.g., SARS-CoV-2 at outbreak) to variants of
concern (e.g., Omicron BA.4/BA.5). We present the results of applying this
approach in an analysis of the U.S. Government / COVID-19 Prevention
Network’s COVE (Moderna) COVID-19 vaccine efficacy clinical trial.
<2022 Schedule>
GOV 3009 Website:
https://projects.iq.harvard.edu/applied.stats.workshop-gov3009
Calendar:
https://calendar.google.com/calendar/embed?src=c_3v93pav9fjkkldrbu9snbhned8…
Best,
Shusei
Dear Applied Statistics Workshop Community,
Our next meeting of the semester will be on September 28 (12:00 EST). Luke
Miratrix and Dae Woong Ham will present "A devil’s bargain? Repairing a
Difference in Differences parallel trends assumption with an initial
matching step."
<Where>
In-person: CGIS K354
Bagged lunches are available for pick-up at 11:40 (CGIS K354).
Zoom:
https://harvard.zoom.us/j/99181972207?pwd=Ykd3ZzVZRnZCSDZqNVpCSURCNnVvQT09
<Abstract>
The Difference in Difference (DiD) estimator is a popular estimator built
on the "parallel trends" assumption that the treatment group, absent
treatment, would change "similarly" to the control group over time. To
increase the plausibility of this assumption, a natural idea is to match
treated and control units prior to a DiD analysis. In this paper, we
characterize the bias of matching under a class of linear structural models
with both observed and unobserved confounders that have time varying
effects. Given this framework, we find that matching on baseline covariates
generally reduces the bias associated with these covariates, when compared
to the original DiD estimator. We further find that additionally matching
on pre-treatment outcomes has both cost and benefit. First, matching on
pre-treatment outcomes will partially balance unobserved confounders, which
mitigates some bias. This reduction is proportional to the outcome's
reliability, a measure of how coupled the outcomes are with the latent
covariates. On the other hand, we find that matching on pre-treatment
outcomes also undermines the second "difference" in a DiD estimate by
forcing the treated and control group's pre-treatment outcomes to be equal.
This injects bias into the final estimate, creating a bias-bias tradeoff.
We extend our bias results to multivariate confounders with multiple
pre-treatment periods and find similar results. We summarize our findings
with heuristic guidelines on whether to match prior to a DiD analysis,
along with a method for roughly estimating the reduction in bias. We
illustrate our guidelines by reanalyzing a recent empirical study that used
matching prior to a DiD analysis to explore the impact of principal
turnover on student achievement.
<2022 Schedule>
GOV 3009 Website:
https://projects.iq.harvard.edu/applied.stats.workshop-gov3009
Calendar:
https://calendar.google.com/calendar/embed?src=c_3v93pav9fjkkldrbu9snbhned8…
Best,
Shusei
Dear Applied Statistics Workshop Community,
Our next meeting of the semester will be on September 21 (12:00 EST).
Matthew Blackwell will present "Difference-in-differences Designs for
Controlled Direct Effects."
<Where>
In-person: CGIS K354
Bagged lunches are available for pick-up at 11:40 (CGIS K354).
Zoom:
https://harvard.zoom.us/j/99181972207?pwd=Ykd3ZzVZRnZCSDZqNVpCSURCNnVvQT09
<Abstract>
Political scientists are increasingly interested in controlled direct
effects, which are important quantities of interest for understanding why,
how, and when causal effects will occur. Unfortunately, their
identification has usually required strong and often unreasonable
selection-on-observables assumptions for the mediator. In this paper, we
show how to identify and estimate controlled direct effects under a
difference-in-differences design where we have measurements of the outcome
and mediator before and after treatment assignment. This design allows us
to weaken the identification assumptions to allow for linear, time-constant
unmeasured confounding between the mediator and the outcome. Furthermore,
we develop a semiparametrically efficient and multiply robust estimator for
these quantities and apply our approach to a recent experiment evaluating
the effectiveness of short conversations at reducing intergroup prejudice.
An open-source software package implements the methodology with a variety
of flexible, machine-learning algorithms to avoid bias from
misspecification.
<2022 Schedule>
GOV 3009 Website:
https://projects.iq.harvard.edu/applied.stats.workshop-gov3009
Calendar:
https://calendar.google.com/calendar/embed?src=c_3v93pav9fjkkldrbu9snbhned8…
Best,
Shusei
Dear Applied Statistics Workshop Community,
Our next meeting of the semester will be on September 14 (12:00 EST). Cory
McCartan will present "Individual and Differential Harm in Redistricting."
<When>
September 14, 12:00 to 1:30 PM, EST
Bagged lunches are available for pick-up at 11:40 (CGIS K354).
<Where>
In-person: CGIS K354
Zoom:
https://harvard.zoom.us/j/99181972207?pwd=Ykd3ZzVZRnZCSDZqNVpCSURCNnVvQT09
<Abstract>
Social scientists have developed dozens of measures for assessing partisan
bias in redistricting. But these measures cannot be easily adapted to other
groups, including those defined by race, class, or geography. Nor are they
applicable to single- or no-party contexts such as local redistricting. To
overcome these limitations, we propose a unified framework of harm for
evaluating the impacts of a districting plan on individual voters and the
groups to which they belong. We consider a voter harmed if their chosen
candidate is not elected under the current plan, but would be under a
different plan. Harm improves on existing measures by both focusing on the
choices of individual voters and directly incorporating counterfactual
plans. We discuss strategies for estimating harm, and demonstrate the
utility of our framework through analyses of partisan gerrymandering in New
Jersey, voting rights litigation in Alabama, and racial dynamics of Boston
City Council elections.
<2022 Schedule>
GOV 3009 Website:
https://projects.iq.harvard.edu/applied.stats.workshop-gov3009
Calendar:
https://calendar.google.com/calendar/embed?src=c_3v93pav9fjkkldrbu9snbhned8…
Best,
Shusei
Dear Applied Statistics Workshop Community,
Our next meeting of the semester will be on September 7 (12:00 EST).
Professor Xiang Zhou presents "Marginal Interventional Effects."
<When>
September 7, 12:00 to 1:30 PM, EST
Bagged lunches are available for pick-up at 11:30 (CGIS K354).
<Where>
In-person: CGIS K354
Zoom:
https://harvard.zoom.us/j/99181972207?pwd=Ykd3ZzVZRnZCSDZqNVpCSURCNnVvQT09
<Abstract>
Abstract: Conventional causal estimands, such as the average treatment
effect (ATE), reflect how the mean outcome in a population or subpopulation
would change if all units received treatment versus control. Real-world
policy changes, however, are often incremental, changing the treatment
status for only a small segment of the population who are at or near “the
margin of participation.” To capture this notion, two parallel lines of
inquiry have developed in economics and in statistics and epidemiology that
define, identify, and estimate what we call interventional effects. In this
article, we bridge these two strands of literature by defining
interventional effect (IE) as the per capita effect of a treatment
intervention on an outcome of interest, and marginal interventional effect
(MIE) as its limit when the size of the intervention approaches zero. The
IE and MIE can be viewed as the unconditional counterparts of the
policy-relevant treatment effect (PRTE) and marginal PRTE (MPRTE) proposed
in the economics literature. However, different from PRTE and MPRTE, IE and
MIE are defined without reference to a latent index model, and, as we show,
can be identified either under unconfoundedness or through the use of
instrumental variables. For both scenarios, we show that MIEs are typically
identified without the strong positivity assumption required of the ATE,
highlight several “stylized interventions” that may be of particular
interest in policy analysis, discuss several parametric and semiparametric
estimation strategies, and illustrate the proposed methods with an
empirical example.
Paper link: https://arxiv.org/abs/2206.10717
<2022 Schedule>
GOV 3009 Website:
https://projects.iq.harvard.edu/applied.stats.workshop-gov3009
Calendar:
https://calendar.google.com/calendar/embed?src=c_3v93pav9fjkkldrbu9snbhned8…
Best,
Shusei
Dear all,
Today the university follows the Monday schedule, so we do not have Applied Stats Workshop.
Our first meeting will be next week (Wednesday, noon). The website (https://projects.iq.harvard.edu/applied.stats.workshop-gov3009) has the Fall schedule.
Best,
Shusei
Dear Applied Statistics Workshop Community,
Our last meeting of the semester will be at *12:10 pm (EST) Wednesday,
April 27*, where Tasha Fairfield
<https://www.lse.ac.uk/international-development/people/tasha-fairfield>
(London
School of Economics and Political Science) presents "Recasting the Debate
on COVID-19 Origins in Bayesian Terms," a joint work with Andrew Charman
(Dept. of Physics, UC Berkeley).
Please note that this meeting will be *entirely on Zoom
<https://harvard.zoom.us/j/97004196610?pwd=eGFydkF5RDRjUlk5RVcyTjV6OStUQT09>*
.
*Abstract*
The debate on covid-19 origins has been politically fraught. Yet setting
aside conspiracy theories and the most implausible of the lab-leak
hypotheses, there is significant disagreement among qualified experts.
Some are adamant that the case should be considered closed in favor of
zoonosis, while others view the evidence as weak, even if they concede that
prior knowledge about previous epidemics favors zoonosis, and a few
maintain that some sort of laboratory leak is a firm possibility.
This project applies the methodology developed in *Social Inquiry and
Bayesian Inference *(CUP 2022) to reassess the debate. We apply Bayesian
reasoning to evaluate the inferential weight of available evidence in favor
of zoonosis vs. lab-leak hypotheses, drawing on published scientific
research, journalistic sources, and interviews with scientists and China
experts. The analysis highlights the flexibility of Bayesian reasoning—this
approach can be used to evaluate any kind of evidence, quantitative or
qualitative, including genetic data, epidemiological data, and information
from interviews and observational fieldwork.
In addition to clarifying the debate by separating prior odds, informed by
what we know from previous epidemics, from the weight of evidence
pertaining directly to SARS-CoV-2, the goals include evaluating to what
extent a Bayesian framework can help improve reasoning when evidence is
limited, communicate degrees of uncertainty more effectively, and
illuminate points of agreement or disagreement among experts on questions
with significant public policy implications.
The table of contents and first chapter of our book are available at:
https://tashafairfield.wixsite.com/home/bayes-book
*Zoom link*:
https://harvard.zoom.us/j/97004196610?pwd=eGFydkF5RDRjUlk5RVcyTjV6OStUQT09
*When:* Wednesday, April 27 at 12:10 - 1:30 pm.
*Schedule of the workshop*:
https://projects.iq.harvard.edu/applied.stats.workshop-gov3009
Looking forward to seeing you all on Wednesday!
Best,
Sooahn
Dear Applied Statistics Workshop Community,
Our next meeting of the semester will be at *12:10 pm (EST) Wednesday, Apr
20*, where Katharina Fellnhofer
<https://sociology.fas.harvard.edu/people/katharina-fellnhofer> (Harvard
University) presents "A framework for measuring intuitive decision making
in real-world contexts."
*Abstract*
Intuition refers to the ability to use nonconscious information for
conscious decision making. The nonconscious element has predominantly been
measured by its speed of operation and ease of application. Only a few
scholarly attempts at behavioral measuring take nonconsciousness into
account, and they use situations that do not represent the real world,
which limits generalization. In my talk, I will present the results of my
intuition measurement using a within-subject design with real investment
opportunities that employ hidden images as nonconscious information to
trigger intuition. My experiments were conducted entirely online from July
to September 2021 in Europe and the United States. I will provide an
overview of my current Bayesian analysis of 62,721 real-world investment
decisions made by 657 subjects representing similar proportions of
financiers, entrepreneurs, and non-entrepreneurs, all recruited via
Prolific. I will also discuss additional ideas that could enrich our
understanding of how to measure skills at using nonconscious information
for conscious, real-world decision making. As such, my presentation will
focus on my existing analytical results, my intentions for future analysis,
and my plans for a new project, with the dual aims of sharing what my team
and I have learned so far and receiving valuable early-stage suggestions
for improvement and feedback from Applied Statistics Workshop participants.
*Where:* CGIS Knafel Building, Room K354
(See this link <https://map.harvard.edu/?bld=04471&level=9> for directions).
*When:* Wednesday, Apr 20 at 12:10 - 1:30 pm.
(The participants can now choose to eat the bagged lunch inside the room
before the presentation starts. You may also pick up the lunch from 11:30
am and eat outside if you wish.)
*Zoom link*:
https://harvard.zoom.us/j/97004196610?pwd=eGFydkF5RDRjUlk5RVcyTjV6OStUQT09
(For the participants who cannot join the session physically.)
*Schedule of the workshop*:
https://projects.iq.harvard.edu/applied.stats.workshop-gov3009
Looking forward to seeing you all on Wednesday!
Best,
Sooahn
Dear Applied Statistics Workshop Community,
Our next meeting of the semester will be at *12:10 pm (EST) Wednesday, Apr
13*, where Deirdre Bloome
<https://www.hks.harvard.edu/faculty/deirdre-bloome> (Harvard University)
presents "Rising Class Crystallization? Trends in Multidimensional Class
Inequality across Racialized/Ethnic Groups."
*Abstract*
In recent decades, U.S. income and wealth inequality grew, educational
attainment rose, and occupational structures shifted. Because these
dimensions of social class are intertwined---with higher education often
generating higher income, wealth, and occupational prestige---rising
inequality in one may have pushed some people toward the tops of multiple
hierarchies, and others toward the bottoms of multiple hierarchies
(polarizing people in the multidimensional space of class inequality). Are
people occupying increasingl*y consistent positions across multiple class
hierarchies*? And has this class *crystallization* trended similarly for
Black, White, and Hispanic people, despite their different opportunities,
constraints, and initial class positions? We address these questions using
data from the Panel Study of Income Dynamics, 1984--2019. To do so, we
introduce nonparametric and parametric methods for studying
multidimensional inequality, including models that jointly parameterize the
mean and covariance matrix of a multivariate outcome as functions of
covariates.
*Where:* CGIS Knafel Building, Room K354
(See this link <https://map.harvard.edu/?bld=04471&level=9> for directions).
*When:* Wednesday, Apr 13 at 12:10 - 1:30 pm.
(The participants can now choose to eat the bagged lunch inside the room
before the presentation starts. You may also pick up the lunch from 11:30
am and eat outside if you wish.)
*Zoom link*:
https://harvard.zoom.us/j/97004196610?pwd=eGFydkF5RDRjUlk5RVcyTjV6OStUQT09
(For the participants who cannot join the session physically.)
*Schedule of the workshop*:
https://projects.iq.harvard.edu/applied.stats.workshop-gov3009
Looking forward to seeing you all on Wednesday!
Best,
Sooahn
Dear Applied Statistics Workshop Community,
Our next meeting of the semester will be at *12:10 pm (EST) Wednesday, Apr
6*, where Adeline Lo <https://www.loadeline.com> (University of
Wisconsin-Madison) presents "Refugees in Modern Media."
*Abstract*
The effects of refugee migration permeates most aspects of a recipient
society, not least native inclusionary attitudes and behaviors towards
refugees. While recent research has emphasized measuring the extent to
which direct exposure to refugees affects inclusion, much less is known
about the more frequent type of refugee exposure natives experience:
exposure to refugees through media representation. This project establishes
key patterns to how much and in what ways modern media represents refugee
stories, how this has changed over time, and explores how major shifts in
the ways refugee stories have changed affect native inclusion using a
unique television corpus covering the universe of broadcasted news in
Germany throughout the period leading up to and following the globally
renowned ``Open Door'' announcement in the Syrian refugee crisis.
*Where:* CGIS Knafel Building, Room K354
(See this link <https://map.harvard.edu/?bld=04471&level=9> for directions).
*When:* Wednesday, Apr 6 at 12:10 - 1:30 pm.
(Bagged lunches available for pick-up at CGIS K354 *11:30 - 11:45 am*, for
the participants who responded to our previous survey. The CGIS cafe on the
first floor has been designated as an eating area, and participants may
also use outdoor spaces for lunch. Please be present at K354 by 12:10 pm
for the presentations.)
*Zoom link*:
https://harvard.zoom.us/j/97004196610?pwd=eGFydkF5RDRjUlk5RVcyTjV6OStUQT09
(For the participants who cannot join the session physically.)
*Schedule of the workshop*:
https://projects.iq.harvard.edu/applied.stats.workshop-gov3009
Looking forward to seeing you all on Wednesday!
Best,
Sooahn
