2005 Colloquium Calendar

Colloquium

MAPSS is proud to announce the inauguration of a new
colloquium series in 2005.

The series has at least four purposes:

1. To bring world-class methodologists from around the world
   to Stanford to give presentations on methodologies of use to social
   scientists across departments at Stanford.
2. To allow Stanford faculty and students to learn more about
   the methodological expertise of our own faculty, who will make
   presentations in the series.
3. To create a sense of community among methodologically
   inclined researchers at Stanford.
4. To provide a weekly yummy and free snack and an interesting
   hour of learning for all members of the Stanford social science
   community.

---

The schedule of presentations in the 2005-2006 series
was:

Date	Loc	Speaker (click name for Bio)	Affiliation	Title (click for Abstract)
2 Mar	McClatchy Hall 215	Katherine Stovel	University of Washington–Professor of Sociology	Hearing About A Job: Networks, Information, and Segregation in Labor Markets
7 Mar	CERAS 204	Simon Jackman	Stanford University–Professor of Political Science	Bayesian analysis in social science settings
14 Mar	S182, GSB	Bradley Efron	Stanford University–Professor of Statistics	50 years of empirical Bayes
11 Apr	Lane History (Bldg 200, Rm 305)	Eric Sheppard	University of Minnesota–Professor of Geography	A methodology for evaluating complex spatial dynamics
18 Apr	Bldg 200, Rm 305	Steve Barley	Stanford University–Professor, Management Science and Engineering	Ethnography
25 Apr	Bldg 200, Rm 305	Leonard Wantchekon	New York University–Professor of Politics	Political field experiments in Africa
2 May	Bldg 200, Rm 305	Tony Bryk	Stanford University–Professor of Education	Hierarchical Linear Modeling
9 May	Bldg 200, Rm 305	J. Scott Long	Indiana University–Professor of Sociology	Comparing groups using predicted probabilities
16 May	Bldg 200, Rm 305	Erik De Corte	Department of Educational Sciences, K.U.Leuven–Professor of Educational Psychology	Design Experiments in Education
23 May	Bldg 200, Rm 305	Helen Longino	Stanford University–Professor of Philosophy	Social epistemology of social science
30 May	Bldg 200, Rm 305	David Freedman	University of California Berkeley -Professor of Statistics	Are experiments better than observational studies? If so, how should we analyze them?

Speaker Bios / Talk Abstracts (as available)

Steve
Barley

Bio: Dr. Stephen R. Barley is
the Charles M. Pigott Professor of Management Science and Engineering
and the Co- Director of the Center for Work, Technology and
Organization at Stanford’s School of Engineering. He holds a Ph.D. in
Organization Studies from the Massachusetts Institute of Technology.
Barley was a member of the Board of Senior Scholars of the National
Center for the Educational Quality of the Workforce and co-chaired
National Research Council and the National Academy of Science’s
committee on the changing occupational structure in the United States.
He has written extensively on the impact of new technologies on work,
the organization of technical work and organizational culture.

Tony
Bryk

Bio: Dr. Anthony Bryk is
Stanfords Spencer Professor of Education, School of Education, and of
Organizational Behavior, Graduate School of Business, and Professor of
Sociology (by courtesy). His long-standing interests are in school
education, educational reform, and educational statistics. He is known
for his work on the Chicago public school system, and particularly for
his development, along with Steve Raudenbush, of methods of and
software for Hierarchical Linear Modeling (HLM) and generalizations
thereof.

Erik
De Corte

Bio: Erik De Corte is Professor
of Educational Psychology at K.U. Leuven in Belgium and currently a
fellow at the Center for Advance Study in Behavioral Science. He is
President of the International Academy of Education and a founder of
the European Association for Research on Learning and Instruction
(EARLI). His research interests include: the learning and teaching of
problem solving skills; metacognition and the affective aspects of
learning; design and evaluation of learning environments; and the
teaching and learning of mathematics.

Bradley
Efron

Bio: Prof. Bradley Efron is the
Max H. Stein Professor of Humanities and Sciences Professor of
Statistics. A member of the National Academy of Sciences, Prof. Efrons
research has focused on the intersection of theoretical and applied
work in statistics, exploring, for example, the underlying statistical
parallels between biostatistics and astrophysics. Creator of the
bootstrap, he has been awarded the Ford Prize, MacArthur Prize, and the
Wilks Medal for his research work in computer applications in
statistics, particularly with techniques known as the bootstrap and the
jackknife. He is largely credited with bringing the power of modern
computers into statistics.

David
Freedman

Abstract: I will suggest that
experiments– when feasible– offer more reliable evidence on causation
than observational studies; this is not to gainsay the contribution to
knowledge from observation. I will also suggest that experiments should
be analyzed as experiments not observational studies. For instance, a
simple comparison of rates might be just the right tool. So far as time
permits, I will discuss models for experimental data, the
intention-to-treat principle, and the effect of treatment on the
treated. I have two related papers on the web, plus a handout:

http://www.stat.berkeley.edu/~census/oxcause.pdf
http://www.stat.berkeley.edu/~census/neyreg.pdf
http://www.stat.berkeley.edu/~census/modelobs.pdf

Bio: David A. Freedman received
his B.Sc. degree from McGill and his Ph.D. from Princeton. He is
professor of statistics at U.C. Berkeley, and a former chairman of the
department. He has been Sloan Professor and Miller Professor, and is
now a member of the American Academy of Arts and Sciences. He has
written several books, including two widely-used elementary texts,
Statistics (with Pisani and Purves, 3rd edition 1997) and Statistical
Models (2005), as well as many papers in probability and statistics. He
has worked on martingale inequalities, Markov processes, de Finettis
theorem, consistency of Bayes estimates, sampling, the bootstrap,
procedures for testing and evaluating models, census adjustment,
epidemiology, cross-level or “ecological” inference, statistics and the
law. In 2003, he received the John J. Carty Award for the Advancement
of Science from the National Academy of Sciences.

Simon
Jackman

Abstract: Bayesian Analysis in Social Science
Settings

What is Bayesian statistical analysis? How does Bayesian
analysis differ from other forms of statistical analysis? Why is the
distinction between Bayesian and non-Bayesian analysis relevant to
statistical work in the social sciences? In this relatively informal
talk, I provide answers to these questions. Via a series of examples
drawn from across the social sciences, I show that a Bayesian approach
offers sensible answers to inferential questions when other approaches
do not or can not. I emphasize the role of advances in computing power,
which has
turned the Bayesian approach from a source of curiosity and controversy
among statisticians and philosophers, into an everyday tool for
empirical researchers in the social sciences. The example applications
include
estimating the partisan bias of an electoral redistricting plan; the
analysis of educational testing data; and tracking public opinion over
an election campaign.

Bio: Prof. Simon Jackman is
Associate Professor, Department of Political Science, and, by courtesy,
Statistics; Director of Graduate Studies in Political Science; and
Director of the Political Science Computational Laboratory. His
principal research and teaching interests are Political Methodology and
American Politics. He has published important articles and lectured
widely on advanced Bayesian methods in econometrics and statistics. A
brilliant communicator of technical and theoretical ideas, he was
awarded the Deans Award for Distinguished Teaching.

J.
Scott Long

Abstract: Social scientists are often
interested in assessing whether the effect of one variable on another
is different in different groups of people. Gender, race, political
party, country, economic region, experimental/control, and class are
but a few examples where group comparisons of the effects of variables
on some outcome are of fundamental importance. Extensions of the Chow
(1960) test are often computed to compare groups, but this approach
confounds the magnitude of the regression coefficients and the error
variances in probit and logit models that assume an underlying latent
variable, often yielding invalid conclusions. Indeed, Allison (1999)
shows that these standard tests confound the magnitude of the
regression coefficients and the variance of the error. Allison proposes
a test that removes the effect of residual variation, but this test
requires auxiliary information that is often unavailable. This lecture
will describe a more feasible approach to comparing group differences
in logit, probit and other types of regression models. We illustrate
how this new method can be used and will describe software that we have
developed to implement the approach.

Bio: Dr. J. Scott Long is
Chancellors Professor, Department of Sociology, Indiana University. He
is on the editorial board of Sociological Methodology, and was awarded
the 2002 Paul F. Lazarsfeld Memorial Award for Distinguished
Contributions to Sociological Methodology. He is well-known for the
important advances he has made in regression models for limited and
categorical dependent variables. Currently, his research examines the
interplay between womens work and family roles and its implications for
physical health; human sexuality and sexual risk-taking, in conjunction
with the Kinsey Institute; and the small sample behavior of robust
(HCCM) standard errors.

Helen
Longino

Bio: Helen Longino, professor
of philosophy, previously was a professor of philosophy and of women’s
studies at the University of Minnesota. Her research interests include
philosophy of science, social epistemology, feminist philosophy and the
development of a social approach to scientific knowledge. Her books
include Science as Social Knowledge (1990) and The Fate of Knowledge
(2002).

Eric
Sheppard

Abstract: A methodology for evaluating complex
spatial dynamics
Eric Sheppard and Paul Plummer

Over the last three decades, considerable effort has been put
into developing methods of analysis for maps: i.e., for spatially
distributed data. These include methods for the identification of
spatial autocorrelation, and spatial econometrics—methods for
rigorous statistical inference in the presence of spatially correlated
data and residuals. Methods also exist for extending time series
analysis to include spatial lags. Yet these methods are of limited
utility for the analysis of the kind of out-of-equilibrium evolutionary
theoretical models that are consistent with complexity theory. Such
models generally are capable of generating a wide range of possible
space-time trajectories, which need to be evaluated in light of a
single observed empirical trajectory. In this talk, we will review
complexity theory and its applicability in economic geography,
summarize the challenges it poses for spatial analysis, and discuss one
proposed methodology for addressing these challenges, based in the
mathematics of coded dynamics.

Bio: Dr. Eric Sheppard is
Professor of Geography at the University of Minnesota, currently a
Fellow at the Center for Advanced Study in the Behavioral Sciences.
Prof. Sheppard is coauthor of The Capitalist Space Economy, a monograph
applying radical political economy as an approach to economic
geography. His current research interests include: How the geographic
nature of society both affects economic change and also requires us to
reconsider the theories we use to understand economic systems;
contested urban futures in North America and Europe; the impact of
information technologies on the geography of globalization; the utility
of GIS for assessing environmental (in)justice.

Katherine
Stovel

Abstract: We present a
framework for simulating labor market matching processes in order to
study mechanisms that generate segregation. Empirical evidence reveals
that labor markets are often highly segregated with respect to the
ascribed attributes of workers. Many occupations are sex-typed, while
in heterogeneous societies certain fields are often dominated by
specific ethnic groups. Various explanations have been proposed to
account for segregation in labor markets. On balance, most of these
explanations can be classified as essentially ’supply-side’ arguments,
emphasizing differences in human capital distributions between groups,
or ‘demand-side’ accounts, based on employer preferences (either
in-group or out-group biases). Yet the process of labor market
stratification is not merely a matter of the human capital
characteristics of workers or the preferences of employers; it also a
function of the complex process by which persons are matched with one
another, including the way that agents in the market find and evaluate
information. In our research we address these oft-overlooked issues by
considering two network-related aspects of the matching process
explicitly: who knows about jobs, and how employers evaluate referrals.
Using an agent-based simulation model, we show how referral hiring
combined with network homophily results in acute segregation by
ascribed attribute, even in the absence of discriminatory preferences
on the part of employers or supply-side human capital differences.
Further, we consider how variation in network structure and network
dynamics can amplify or mitigate these effects.

Bio: Katherine Stovel (A.B
Stanford, Political Science; PhD. UNC-Chapel Hill, Sociology) is
Associate Professor of Sociology at the University of Washington. While
her research spans number of substantive areas–including economic
sociology, adolescent health, and quantitative methods–her work is
motivated by a general concern with how basic principles of social
interaction are expressed in specific historical or cultural contexts,
and why these expressions may result in new institutional arrangements
or new identities for individuals.

Leonard
Wantchekon

Abstract: This paper provides
experimental estimates of the effect of ethnic ties between voters and
candidates on electoral support for nation-building policies. The
estimates are based on voting outcomes in selected non-competitive
districts that were randomly assigned to
“purified” national public goods and redistributive platforms by
candidates competing in the 2001 presidential elections in Benin. We
find that ethnic ties do not weaken support for nation public goods
platforms. The effect is even positive and significant in some cases,
especially among women. In addition, we find that the positive effect
of
ethnic ties would have remained, even if the experiment has taken place
in more competitive and ethnically diverse districts. The results
suggest that ethnic ties can help secure electoral support for
nation-building policies.

Bio: Dr. Leonard Wantchekon is
Associate Professor of the Wilf Family Department of Politics at New
York University. He is founder and director of the Institute for
Empirical Research in Political Economy. His research interests include
Political Economy, Development, Applied Game Theory and Comparative
Politics. He has published in the American Political Science Review,
Comparative Political Studies, and the Quarterly Journal of Economics,
among others.