Hi all,
We hope you can join us for the Applied Statistics Workshop this
Wednesday, September 29th, when we are excited to have Jamie Robins
speaking on direct and indirect effects, based on joint work with
Thomas Richardson. Details of his talk, an abstract, and a link to the
paper are below. The workshop begins at 12 noon and we will serve a
light lunch. Hope to see you there!
"Pure (natural) direct and indirect effects, alternative
counterfactual models, and the philosophy of science"
James Robins
Harvard School of Public Health
September 29th, 2010, 12 noon
K354 CGIS Knafel (1737 Cambridge St)
Paper: http://www.iq.harvard.edu/events/sites/iq.harvard.edu.events/files/wp100.pdf
Abstract:
We consider four classes of graphical causal models: the Finest Fully
Randomized Causally Interpretable Structured Tree Graph (FFRCISTG) of
Robins (1986), the agnostic causal model of Spirtes et al. (1993), the
Non-Parametric Structural Equation Model (NPSEM) of Pearl (2000), and
the Minimal Counterfactual Model (MCM) which we introduce. The latter
is referred to as ‘minimal’ because it imposes the minimal
counterfactual independence assumptions required to identify those
causal contrasts representing the effect of an ideal intervention on
any subset of the variables in the graph. The causal contrasts
identified by an MCM are, in general, a strict subset of those
identified by a NPSEM associated with the same graph. We analyze
various measures of the ‘direct’ causal effect, focusing on the pure
direct effect (PDE), also called the ‘natural direct effect’. We show
the PDE is a parameter that may be identified in a DAG viewed as a
NPSEM, but not as an MCM. In spite of this, Pearl has given a scenario
in which the PDE corresponds to the intent-to-treat parameter of a
randomized experiment. We resolve this apparent paradox by showing
that implicit within Pearl’s account is an extended causal DAG with
additional variables in which there is a causal contrast that equals
the pure direct effect of Pearl’s original NPSEM. Further, this
contrast is identified from observational data on the original
variables. Finally we relate our results to the work of Avin et al.
(2005) on path-specific causal effects.
~~~~~~~~~~~
Matthew Blackwell
PhD Candidate
Institute for Quantitative Social Science
Department of Government
Harvard University
url: http://people.fas.harvard.edu/~blackwel/
Hi all,
We hope you can join us for the Applied Statistics Workshop of the
year this Wednesday, September 22nd, when we will be happy to have
César A. Hidalgo presenting a paper entitled "The Structure of
Economic Complexity." César is an Assistant Professor at the MIT Media
Lab, where he leads the Macro Connections group, and a faculty
associate at Harvard University’s Center for International Development
Details of his talk and an abstract are below. The workshop begins at
12 noon and we will serve a light lunch. Hope to see you there!
"The Structure of Economic Complexity"
César A. Hidalgo
Assistant Professor, Massachusetts Institute of Technology (MIT)
September 22nd, 2010, 12 noon
K354 CGIS Knafel (1737 Cambridge St)
Abstract:
With billions or products, billions of people, and trillions of
interactions, the world economy is one of the most outstanding complex
systems to have ever emerged. Can we use complexity science to improve
our understanding of a system of such paramount complexity? In this
talk I summarize recent research that uses networks to describe,
characterize and understand differences in the productive structure of
nations and their evolution. First, I show how the complexity of an
economy can be quantified by looking at the structure of the network
connecting countries to the products they export and that countries
tend to approach a level of income which is dictated by the complexity
of their economies. Then, I show how development is constrained by a
projection of this network into the space of products, or Product
Space, by demonstrating empirically that the evolution of countries
comparative advantage is constrained by the structure of this network.
To conclude I present a simple model that can account for some of the
stylized facts that we uncover from our analysis and discuss how
strategy and policy can be reinterpreted from this alternative
perspective.
Cheers,
matt.
~~~~~~~~~~~
Matthew Blackwell
PhD Candidate
Institute for Quantitative Social Science
Department of Government
Harvard University
url: http://people.fas.harvard.edu/~blackwel/
Hi all,
We hope you can join us for the Applied Statistics Workshop this
Wednesday, September 15th, when I will have the dubious honor of
hosting myself, Matthew Blackwell, and presenting joint work with
James Honaker and Gary King. Details of the talk and an abstract are
below. The workshop begins at 12 noon and we will serve a light lunch.
Hope to see you there!
"Multiple Overimputation: A Unified Approach to Measurement Error and
Missing Data"
Matthew Blackwell (with James Honaker and Gary King)
Governemnt
September 15th, 2010, 12 noon
K354 CGIS Knafel (1737 Cambridge St)
Abstract:
Social scientists typically devote considerable effort to reducing
measurement error during data collection only to ignore the issue
during data analysis. Although many statistical methods have been
proposed for reducing measurement error-induced biases, few have been
widely used because of implausible assumptions, high levels of model
dependence, difficult computation, or inapplicability with multiple
mismeasured variables. We develop an easy-to-use alternative that
generalizes the popular multiple imputation (MI) framework so that it
treats missing data problems as a special case of extreme measurement
error and corrects for both. Like MI, the proposed "multiple
overimputation" (MO) framework is a simple two-step procedure. First,
multiple (~5) completed copies of the dataset are created where cells
measured without error are held constant, those missing are imputed
from the distribution of predicted values, and cells (or entire
variables) with measurement error are "overimputed," that is imputed
from a predictive distribution with observation-level priors defined
by the mismeasured values and available external information, if any.
In the second step, analysts can then run whatever statistical method
they would have run on each of the overimputed datasets as if there
had been no missingness or measurement error; the results are then
combined via a simple procedure. We also offer open source software
that implements all the methods described herein.
You can find a copy of the paper here:
http://gking.harvard.edu/files/abs/measure-abs.shtml
Cheers,
matt.
~~~~~~~~~~~
Matthew Blackwell
PhD Candidate
Institute for Quantitative Social Science
Department of Government
Harvard University
url: http://people.fas.harvard.edu/~blackwel/
Hi all,
We hope you can join us for the first Applied Statistics Workshop of
the year this Wednesday, September 8th, when we will be happy to have
Brigham Fransden who is a Robert Wood Johnson Scholar in Health
Policy. Brigham received his PhD from the Department of Economics at
MIT this past June. Details of his talk and an abstract are below. The
workshop begins at 12 noon and we will serve a light lunch. Hope to
see you there!
"Nonparametric randomization inference on quantiles under imperfect compliance"
Brigham Fransden
Robert Wood Johnson Scholar in Health Policy
September 8th, 2010, 12 noon
K354 CGIS Knafel (1737 Cambridge St)
Abstract:
I develop a nonparametric approach to randomization inference on the
effects of a treatment under imperfect compliance to a randomized
assignment. I show how to construct exact finite-sample confidence
intervals for the distribution of potential outcomes for the subgroup
of subjects---compliers---whose treatment status is manipulated by the
random assignment. An advantage of this approach is that it allows the
effect of treatment to be heterogeneous and makes inference on the
effects of the treatment across the distribution without imposing a
parametric model.
Cheers,
matt.
~~~~~~~~~~~
Matthew Blackwell
PhD Candidate
Institute for Quantitative Social Science
Department of Government
Harvard University
url: http://people.fas.harvard.edu/~blackwel/