Sharing a new version of the paper "Policy Design in Experiments with Unknown Interference" that studies experimental design with spillover effects. The paper now incudes an implementation of the proposed design in a large scale experiment in Pakistan:
I am excited to announce that I will join
@HarvardEcon
as an assistant professor! Before that, I will spend one year as a postdoc at
@StanfordGSB
. I am grateful to my advisors,
@UCSDEcon
, family, friends, and many people met these years for these achievements
Happy to announce that my paper: "Policy Targeting under Network Interference" has just been accepted at
@RevEconStudies
The paper studies how to optimally allocate treatments in the presence of spillover effects, using info from an existing experiment or obs/ study 1/3
Many thanks to the community at
@StanfordGSB
@StanfordEcon
for this postdoc year, and special thanks to
@guido_imbens
and the metrics group for the mentorship, it was a truly incredible experience. Ready and excited to join the faculty at
@HarvardEcon
starting this summer!
New working paper with a fantastic team of coauthors:
we study when and how to design cluster experiments with network spillovers.
with
@lihua_lei_stat
,
@guido_imbens
, Brian Karrer, Okke Schrijvers, Liang Shi
``One challenge in experiments or obs/ studies is that treating an individual may generate spillovers, affecting the design of the welfare-maximizing policy. Here, I present methods to optimally allocate treatments with spillovers."
Recently accepted paper from
@VivianoDavide
👇
Many thanks to the American Statistical Association - Business and Economic Statistics Section - for the Zellner Thesis Award in Econometrics and Statistics for my PhD thesis. More details here
Postdoctoral fellowship alert:
The Harvard Data Science Institute is hiring postdoctoral fellows in Data Science. Applicants must specify two to five Harvard faculty to work with from *any* department -- Deadline is November 13, apply!
Excited to announce that I am on the job market. My research combines economics and data science to develop or justify methods for applications in econ/social sciences. My JMP studies experimental designs with spillover effects.
For more info visit 1/3
REStud Tours, initiated in 1989, have since embodied REStud’s tradition of promoting original studies by young economists.
Here are the selected speakers for 2022 and summaries of their promising JMPs. 1/8
🧵
Happy to share the new version of our wp on multiple hypothesis testing joint with
@wuthrich_k
and
@PaulFNiehaus
, which now also include an empirical analysis of research costs for MHT interpreted through our model.
Congratulations Dr Viviano
@VivianoDavide
after a successful defense. It took a whole team of advisors (not all in photo) to keep up with your work. Good luck at
@StanfordGSB
for the postdoc and then
@HarvardEcon
.
ECMA is happy to announce the 2024 Frisch Medal has been awarded to Dennis Egger, Johannes Haushofer, Edward Miguel, Paul Niehaus, & Michael Walker for their outstanding paper: General equilibrium effects of cash transfers: experimental evidence from Kenya
Hi
#EconTwitter
!✍️
Working on perfecting your writing style for
#economics
papers and essays?
Check out this useful resource by
@VaranyaChaubey
on improving your research paper writing skills. Tailored for PhD candidates and postdocs but useful for everyone!
Plus, don’t miss
#ASSA2023
will feature many interesting sessions on experimental design, causal inference, and policy learning. For those interested, I will be presenting on Jan 6, session on Adaptive Experimentation and Policy Learning ().
Excited to present here our work that studies when and how to adjust inference for multiple hypothesis testing ( , w/ Kaspar Wuthrich and
@PaulFNiehaus
)
The Workshop in Applied Statistics kicks off the new academic year with
@VivianoDavide
(Harvard Department of Economics), who is presenting "Policy targeting under interference." Join us at 12:00 noon!
Thanks
@NUEconomics
and
@UChi_Economics
for wonderful visits this week, and looking forward to the interaction conference on econometrics and applied Econ at
@WIeconomics
this weekend!
Amazing network economics workshop is coming up in the Twin Cities
@UMNISyE
with
@VivianoDavide
@ben_golub
Matt Jackson, and many more.
And,
@korenmiklos
will present our paper on collaboration in open-source software.
If instead you are interested in statistical treatment choice with networks/spillovers using data from an existing (quasi)experiment, I also recently posted a revised version on my other paper on choosing whom to treat with spillover effects here:
An econometrician, a macro-economist and a psychologist walk into a Yale conference room.... And end to a day with robust discussion about harmonizing outcomes, thinking through heterogenity, scaling questions. More to come! n/n
If you are interested in networks, causal inference, and experimental design, I will be presenting my works in two separate sections at the EEA-ESEM congress this week (Tuesday and Wednesday). Details are here
@stats_tipton
If it can be helpful, in this paper we argue that multiple hypothesis adjustments should mostly depend on economic incentives and costs complementarities of each study.
In his paper “Policy Design in Experiments with Unknown Inference,”
@GSBSilab
research affiliate
@VivianoDavide
explores experimental design with spillover effects. In its latest version, he implements it on a large-scale experiment in Pakistan
The method extends the lit on policy learning to the context of spillovers, allowing for local spillovers in the network and constraints on the policy space.
For implementation, see the R-package NetworkTargeting developed with
@J_S_Carlson
“They bring us here and let us construct nice ships where people can go on vacation,” says Sani who migrated from Bangladesh at age 16 to join his father, a shipyard worker. “But when they have to recognise your rights, they want you to be a slave.”
BREAKING NEWS:
The 2021 Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel has been awarded with one half to David Card and the other half jointly to Joshua D. Angrist and Guido W. Imbens.
#NobelPrize
My research agenda includes a) experimental design and inference in experiments; b) policy design and inference with quasi-experiments, studied in the contexts of (i) spillover effects/networks, (ii) machine-learning for causal inference, and (iii) algorithmic fairness. 2/3
By leveraging a sampling randomization device with information from neighbors of (quasi)experiment participants, I identify and estimate the target welfare. I then solve an optimization program and characterize its theoretical properties
Hello folks, here is my first (presentational) tweet: I am a Ph.D. candidate at UC San Diego, Econ working on econometrics and Data Science, with a focus on causal inference, networks, experimental and policy design. You find more info at
Last but not least many (many) thanks to the Editor (Francesca Molinari) and all four referees who did a terrific job which substantially improved the paper. And very special thanks also to my advisors and colleagues for incredibile support and feedback
@AlejandroSBe
@clintmharris1
Interestingly whether and how we want to do indexing vs betw/ outcomes multiple hypothesis testing corrections depends on how we interpret such outcomes from a decision theoretic perspective. We have some discussion in Sec 4 here in case of interest
@Apoorva__Lal
@mjskay
We have some brief comments on FDR in Sec 3, although your setting seems different from ours. The Bayesian interpretation of pFDR through q-values in Storey's 2003 paper might help in your case. ()
(Southern) Italians will understand
The list in the center is the menu. The other notes split it by cook. Note that Vittorio is my grandpa (90 y.o. next April).
@MichelaTincani
@m_maggiori
I think studying latin (and ancient greek) was somehow useful to learn how to learn, but agree that studying other subjects instead might be more useful.
@paulgp
@RevEconStudies
Indeed I do! it took me some time for the jmp as there we waited to then implement it in a real large scale experiment after the job market