Current Projects and Working Papers
All titles are preliminary and comments are welcome.
Mouse over titles for abstracts when available.
Applied Statistics
Randomization Inference and/or Matching for Causal and Statistical Inference
"The randomization mode of statistical inference" A perspective on randomization inference in experiments with a methodological proposal for the use of regression in adjustment for experiments. (with Costas Panagopoulos).
Making Effects Manifest in Randomized Experiments Experimentalists desire precise estimates of treatment effects and nearly always care about how treatment effects may differ across subgroups. After data collection, concern may focus on random imbalance between treatment groups on substantively important variables. Pursuit of these three goals --- enhanced precision, understanding treatment effect heterogeneity, and imbalance adjustment --- requires background information about experimental units. For example, one may group similar observations on the basis of such variables and then assign treatment within those blocks. Use of covariates after data have been collected raises extra concerns and requires special justification. For example standard regression tables only approximate the statistical inference that experimentalists desire. The standard linear model may also mislead via extrapolation. After providing some general background about how covariates may, in principle, enable pursuit of precision and statistical adjustment, this paper presents two alternative approaches to covariance adjustment: one using modern matching techniques and another using the linear model --- both use randomization as the basis for statistical inference. (to appear in the Cambridge Handbook of Experimental Political Science). The compendium, or reproduction archive, for the paper is available as a package for the R statistical computing environment. See this little tutorial file for detailed instructions about how to download and use it.
"Probability of What?": A Randomization-based Method for Hypothesis Tests and Confidence Intervals about Treatment Effects This drafty working paper provides my perspective on randomization-based inference for randomized experiments. (with Costas Panagopoulos).
Attributing Effects to A Cluster Randomized Get-Out-The-Vote Campaign. (with Ben Hansen ). University of Michigan, Dept of Statistics Technical Report # 448. October 2006. (pdf version here)
Attributing Effects to A Cluster Randomized Get-Out-The-Vote Campaign: An Application of Randomization Inference Using Full Matching
Statistical analysis requires a probability model: commonly, a model for the
dependence of outcomes Y on confounders X and a potentially causal variable
Z. When the goal of the analysis is to infer Z’s effects on Y, this requirement
introduces an element of circularity: in order to decide how Z affects Y, the
analyst first determines, speculatively, the manner of Y ’s dependence on Z and
other variables. This paper takes a statistical perspective that avoids such cir-
cles, permitting analysis of Z’s effects on Y even as the statistician remains
entirely agnostic about the conditional distribution of Y given X and Z, or
perhaps even denies that such a distribution exists. Our assumptions instead
pertain to the conditional distribution Z|X, and the role of speculation in set-
tling them is reduced by the existence of random assignment of Z in a field
experiment as well as by poststratification, testing for overt bias before accept-
ing a poststratification, and optimal full matching. Such beginnings pave the
way for “randomization inference”, an approach which, despite a long history in
the analysis of designed experiments, is relatively new to political science and
to other fields in which experimental data are rarely available.
The approach applies to both experiments and observational studies. We
illustrate this by applying it to analyze A. Gerber and D. Green’s New Haven
Vote 98 campaign. Conceived as both a get-out-the-vote campaign and a field
experiment in political participation, the study assigned households to treat-
ment and desired to estimate the effect of treatment on the individuals nested
within the households. We estimate the number of voters who would not have
voted had the campaign not prompted them to — that is, the total number of
votes attributable to the interventions of the campaigners — while taking into
account the non-independence of observations within households, non-random
compliance, and missing responses. Both our statistical inferences about these
attributable effects and the stratification and matching that precede them rely
on quite recent developments from statistics; our matching, in particular, has
novel features of potentially wide applicability. Our broad findings resemble
those of the original analysis by Gerber and Green (2000).
(with Ben Hansen ) prepared for presentation at the Political Methodology meetings, July 2005
Attributing Effects to a Get-Out-The-Vote Campaign Using Full Matching and Randomization Inference (with Ben Hansen ) prepared for presentation at the MPSA meetings, April 2005
Miscellaneous Fun Stuff
Cycling Involvements: Frequency Domain Time Series Analysis and Political Participation in the USA This paper shows that decomposing a time-series into periodic components can provide po- litically useful information about the shape of aggregate political participation in the United States. Specifically, it provides statistical tests for the periodicity of the aggregate time series of political participation and explains how this decomposition and associated tests work. Between 1973 and 1994 there appears to be an annual cycle in the reporting of political participation by respondents to a series of polls conducted by Gallup 10 times per year. This seasonality has been noted by in one other publication, by Rosenstone and Hansen (1993), but was explained as tied to a summer political cycle. In this article I suggest that this discovery has more to do with annual cycles in the composition of the Gallup sample than politics. I am currently trying to obtain detailed information on the monthly mail volume into and out of Congress. With this information, I will be able to test more directly if, despite the changes in sample composition of the Gallup polls, the political participation of Americans ought to be see as an "output" of Congressional mobilization or an "input" or in what way the flow of participation into Congress is related to the flow of mobilization out of it. Very drafty and still waiting for data on monthly mail to and from Congress. This paper contains a basic description of some frequency domain time series analysis as applied to a political science topic.
Political Behavior
The Shape of Political Participation
Although 50 years of excellent scholarship have taught us a great
deal about how political participation varies across individuals
within one point in time, scholars do not know much about how
political participation changes over time within the lives of
individuals. By focusing predominantly on the preconditions
for participation, the literature has largely ignored the
precipitants of it. In this paper, I
endeavor to show what political participation looks like if we think
of it as a process evolving year-by-year across the lives of
ordinary people. The new description offered here provides some
evidence that challenges the basis for extant theories of why
individuals participate in politics. The purpose of this paper is
not to offer new theories or frameworks for understanding, but
merely to offer a new vision of what political participation
is; a vision which differs from and, I hope, complements that
currently assumed by scholars in this field; a vision that, I hope,
spurs new theories and new modes of research in this area.
A Framework for Studying the Dynamics of Political Participation.
Political action is driven by events. Although the effects of
particularly dramatic events on social movements is well documented,
the effects of events, quotidian or exceptional, on the behavior
of individuals are significantly less well understood. This paper
proposes a framework for understanding how a moment of political
action may occur in the life of an ordinary person. It synthesizes
past literature and theories that explain variation among people at a
single point in time on the basis of largely time-constant
attributes of people and elaborates on this literature to suggest
when we might expect the poor and disadvantaged to surmount such
resource, skills, and status barriers to get involved in
politics. Furthermore, this framework suggests a way for future
syntheses, theory-building, and empirical studies to coordinate such
that all of our disparate findings about political participation
cumulate more effectively.
Civic engagement of the World War II generation. (with Suzanne Mettler and Theda Skocpol)
Words to Math: Representing Theoretical Concepts Mathematically --- The Case of "Symbolic Racism" (with Cara Wong )
The True Story of the Literary Digest Poll of 1936 (with Adam Berinsky )
Threat, Mobilization, and Participation: The Impact of Crossburnings on Political Behavior in North Carolina
Black Threat, Christian Threat, Black Context, Christian Context
Unpacking Power Threat Theory (with Cara Wong )