Colloquium

MAPSS is proud to announce the inauguration of a new
colloquium series in 2006.

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.

Date	Loc	Speaker (click name for Bio)	Affiliation	Title (click for Abstract)
10 Oct	Education Room 313	Persi Diaconis	Stanford University Professor of Statistics and Mathematics	“Horse Shoes and Politics Do Mix”
07 Nov	Education Room 313	Dave MacKinnon	Arizona State University Professor of Psychology	Designs for Assessing Mediating Variables
14 Nov	Bldg. 200, Room 305	Helena Chmura Kraemer and Michaela Kiernan	Stanford University Professor of Biostatistics in Psychiatry (Kraemer) and Senior Research Scientist, Stanford Prevention Research Center (Kiernan), School of Medicine	Moderators and Mediators: Comparing the Baron & Kenny and MacArthur Approaches (link to PowerPoint file, 185k)
03 Dec	Hewlett 201	Edward R. Tufte	Yale University Professor Emeritus of Political Science, Computer Science, and Statistics, Senior Critic, School of Art.	An Academic and Otherwise Life, an N = 1
05 Dec	Bldg. 200, Room 305	Douglas Rivers	Stanford University Professor of Political Science & Senior Fellow, the Hoover Institution	Sampling for Web Surveys
Winter Term
16 Jan	Bldg. 550, Room 550A	Ken Bollen	University of North Carolina Professor of Sociology and Statistics	Structural Equation Models (SEMs) for Pooled Time-Series and Cross Sectional (Panal) Data
06 Feb	Bldg. 550, Room 550A	Shanto Iyengar	Stanford University Professor of Communication and Political Science	Experimental Designs for Political Communication Research: From Shopping Malls to the Internet
13 Feb	Bldg. 550, Room 550A	Don Green	Yale University Professor of Political Science	Gauging the Influence of Political Advertising on Television and Radio: Results from a Large Scale Randomized Experiment (link to associated paper)
20 Feb	Bldg. 550, Room 550A	Robert N. Parker	University of California, Riverside, Professor of Sociology, Co-Director of Presley Center for Crime and Justice Studies	GIS and Social Research: Data Dissemination, Database Construction, and Multivariate Analysis
06 Mar	Bldg. 550, Room 550A	Sean Reardon	Stanford University Professor of Sociology and Education	Propensity Score Matching for Causal Inference: Possibilities, Limitations, and an Example (link to presentation slides)
Spring Term
24 Apr	Bldg. 200, Room 203	Jay McClelland	Stanford University Professor of Psychology, and Director, Center for Mind, Brain and Computation	Dynamical Models of Decision Making: Optimality, human performance, and principles of neural information processing (link to presentation slides)
08 May	Bldg. 200, Room 203	Jim Moody	Duke University Professor of Sociology	Diffusion over Dynamic Networks
15 May	Bldg. 200, Room 203	Linda Piekarski	Vice President for Database and Research, Survey Sampling International	Surveys in the 21st Century
22 May	Bldg. 200, Room 203	Christian Wheeler	Stanford University Graduate School of Business Professor of Marketing	Manipulation and Measurement of Construct Accessibility
29 May	Bldg. 200, Room 203	Paul Sniderman and Michael Tomz	Stanford University Professors of Political Science	The Microfoundations of Issue Voting
05 June	Bldg. 200, Room 203	Mollyann Brodie	Director, Public Media and Media Research, Kaiser Family Foundation	Using Public Opinion Research to inform Public Policy Debates in ‘Real Time’

All members of the Stanford community are invited to attend
and to RSVP in advance to reserve a place.
To RSVP, please click here.

Lunch will be served at 11:45 for those who have
RSVP’d;
the talks start at noon.

For more information, please contact mapss-info@lists.stanford.edu.

Speaker Bios / Talk Abstracts (as available)

Persi
Diaconis

“Horse Shoes and Politics Do Mix”
Abstract: Many data displays aimed at reducing dimensionality result in
an “artifact” called a horseshoe. In joint work with Sharad Goel and
Susan Holmes, we have found an explanation and an interpretation. This
allows detection of gradients (seriation). The ideas are applied to the
current voting data in the US House of Representatives.

Bio: Dr. Persi Diaconis is Mary V. Sunseri Professor of
Statistics and Mathematics. He is known for tackling mathematical
problems involving randomness and randomization,
such as coin
flipping and shuffling
playing cards. In his early teens Dr. Diaconis dropped out of
Julliard, where he studied violin, to become a professional magician,
which he was for eight years. He returned to school to study
probability and gained his PhD in Mathematical Statistics from Harvard
in 1974. He has published widely in mathematics and statistics, and has
served as statistical consultant to many companies and research labs.
In 1982 he was awarded a five-year MacArthur Fellowship, or “Genius
Grant,” and in 1995 was made a member of the National Academy of
Sciences.

Dave
MacKinnon

“Designs for Assessing Mediating
Variables”
Abstract: Among the single hottest social science methodology
innovations of the last century is the testing of mediation.
Almost every social science hypothesis proposes a mechanism by which
one variable affects another. A “mediating variable” is that mechanism.
For example, if having almost no money prevents people from
buying food, and being hungry causes people to become depressed, then
ability to buy food is a mediator by which income influences
depression. Mediation is at the core of many social science
theories: attitudes cause intentions, which in turn cause behavior;
poverty reduces local social ties which increases crime rates; parent
characteristics determine a child’s environment, which determines the
child’s achievement. The most popular approach to testing mediational
hypotheses involves analyzing correlations among variables and entails
making a series of usually untestable assumptions. This talk
will describe alternative ways to test mediation through
experimentation: designs in which a researcher randomly assigns
participants to levels of the independent and mediating variables,
designs that selectively block or enhance a mediational relation, and
sequences of studies that test whether the pattern of results matches
the hypothesized mediation relation.

Bio: Dr. David MacKinnon is one of the nation’s leading
experts on testing mediation. He is a quantitative
psychologist interested in (1) developing new statistical methods in
social science research, and (2) prevention research and health
psychology. He graduated from Harvard University in 1979 and received
the Ph.D. degree in Measurement and Psychometrics from UCLA in 1986.
He was a professor at the University of Southern California
before joining the Arizona State University faculty, where he is
Professor of Psychology and director of the Research in Prevention
Laboratory at Arizona State University. He teaches
graduate and undergraduate research methods and statistics courses and
has received continuous federal funding of his research since 1990.

Helena
Chmura Kraemer and Michaela Kiernan

“Moderators and Mediators:
Comparing the Baron & Kenny and MacArthur Approaches”
Abstract: This is our second talk in the series on mediation, and this
one expands the scope to address moderation as well, which is the
process by which one variable changes the effect of a second variable
on a third variable. Establishing moderators and mediators
to elucidate causal chains is essential for social science research,
and the most popular set of statistical tests used to do this were
outlined by Baron and Kenny in 1986. Recently, a new set of
statistical testing methods were proposed by a MacArthur network
subgroup to extend the Baron and Kenny approach and resolve its
ambiguities by: (1) imposing screening criteria that identify whether a
variable is eligible for consideration as a moderator or mediator; (2)
imposing analytic criteria that demonstrate whether an eligible
variable functions as a moderator or mediator; (3) not drawing causal
inferences from observational data; and (4) considering other relations
among variables besides moderation and mediation. To assist
researchers considering moderator/mediator issues in their choice of
which approach is better suited to their research needs, we will
explain these various approaches.

Bios: Dr. Helena Chmura Kraemer is Professor of Biostatistics
in Psychiatry in the Department of Psychiatry and Behavioral Sciences
of Stanford University. Dr. Kraemer’s research focuses on experimental
and observational research methods in the behavioral aspects of
medicine. Her work has mostly been in psychiatry and health
psychology, but has also touched on other fields of medicine, including
cardiology, epidemiology, pediatrics, and oncology. Her most
recent work has focused on effect sizes that convey clinical
significance and on moderators and mediators of treatment in randomized
clinical trials.

Edward
R. Tufte

“An Academic and Otherwise Life, an N = 1
”

Abstract: The ‘da Vinci of data,’ (New York Times, 1998), Edward Tufte
is the author of books on graphic presentation of data that have been
an inspiration for designers, information architects, engineers, and
scientists around the world. He is the world’s leading authority on the
design of visual data displays. Winner of the 2004 American Institute
of Graphic Arts medal and dozens of other awards, Tufte began his
career as a world-renowed political scientists and statistician and
transformed himself mid-career into a leader of analytical design.
During this colloquium for Stanford students and faculty, Tufte will
describe the progress of his life and career through its many
fascinating stages.

Douglas
Rivers

“Sampling for Web Surveys”
Abstract: With steeply increasing costs of conducting
face-to-face and telephone interviews for surveys of representative
samples of people and declines in response rates in such surveys, the
appearance of the Internet as a way to conduct surveys has tremendous
appeal. When a researcher has a list of the entire
population of interest and email addresses for them all, a survey of a
representative sample is easy to conduct defensibly. But
researchers are often interested in doing Internet surveys of a
representative sample of the general population, for which no complete
list exists. Challenges entailed include the fact that some
members of the population do not have Internet access at all, even
people who do have internet access are not listed in any way that would
allow drawing a probability sample, Internet surveys sometimes elicit
very low response rates, and respondents sometimes choose not to answer
many questions. This talk will review methods for overcoming
these challenges and will make comparisons to random digit dial
telephone surveys.

Bio: Dr. Douglas Rivers is one of the
world’s leading experts on survey research and a successful Silicon
Valley entrepreneur. Before joining the Stanford faculty, he taught at
Harvard University, Caltech, and UCLA. At Stanford, he is
Professor of Political Science and Senior Fellow at the Hoover
Institution. Rivers has founded three successful technology companies,
Preview Systems, Knowledge Networks, and Polimetrix. He has published
numerous academic papers in such outlets as the American Political
Science Review, American Journal of Political Science, Econometrics
Journal, and the Journal of Econometrics, and is a consultant to CBS
News. He holds a B.A. from Columbia University and a Ph.D. from Harvard
University.

Ken Bollen

“Structural Equation Models (SEMs)
for Pooled Time-Series and
Cross Sectional (Panel) Data”
Abstract: Econometrics and statistics have developed a variety
of
models and estimators that take advantage of the special
characteristics of
pooled time-series and cross-sectional data. Also known as “fixed” or
“random” effect estimators for panel or
longitudinal data, there are numerous applications of these techniques
in the
social science literature. The purpose of this paper is
to illustrate the
integration of these techniques into a structural equation model (SEM)
framework that permits easy estimation of standard and nonstandard
variants of
the fixed and random effects model. I
discuss likelihood ratio tests of fixed vs. random effects model and
equal vs.
unequal coefficients of the time-varying variables, including
time-invariant
variables in fixed effect models, treating missing data, and other
extensions. By
integrating these pooled
time-series and cross-sectional techniques into SEMs, we can take
advantage of
the estimation, testing, and fit assessment capabilities that are
readily
available for SEMs.

Bio: Kenneth A. Bollen is Director of the Odum Institute for
Research in Social Science and the H.R. Immerwahr Distinguished
Professor of Sociology at the University of North Carolina at Chapel
Hill. He is the Year 2000 recipient of the Lazarsfeld Award for
Methodological Contributions in Sociology. The ISI named him among the
World’s Most Cited Authors in the Social Sciences. He is
coauthor of Latent Curve Models: A Structural Equations Approach (with
P. Curran, 2006, Wiley) and author of Structural Equation Models with
Latent Variables (1989, Wiley) and of over 100 papers.
Bollen’s primary areas of statistical research are structural
equation
models and latent curve models.

Shanto
Iyengar

“Experimental Designs for Political Communication
Research: From Shopping Malls to the Internet”
Abstract: I argue that field methods and the technological advances
associated with the Internet have gone a long way towards neutralizing
the traditional weaknesses of experimentation. First,
experiments administered online can prove just as realistic and
generalizable as conventional experiments. Second, the
greater reach of online experiments permits a more heterogeneous
subject pool. Moreover, sampling bias can be eliminated
entirely through the use of standard probability sampling techniques to
the recruitment of experimental participants. Third,
experimental design can be adapted to permit voluntary exposure to the
stimulus, thus reflecting the inherent selectivity in the composition
of real-world media audiences. Thus, experiments now represent a
dominant methodology for mass communication research.

Bio: Shanto Iyengar is the Harry and Norman Chandler Professor
of Communication at Stanford University. He also has a joint
faculty position in Political Science and is the director of the
Political Communication Lab, which is a research group that utilizes
the Internet to study politics and media. His research has been
published in several journals including American Political Science
Review, Communication Research, Journal of Personality and Social
Psychology, and Public Opinion Quarterly. Iyengar’s books
include Going Negative: How Political Advertisements Shrink and
Polarize the Electorate (co-authored with Steven Ansolabehere), Do the
Media Govern? (co-edited with Richard Reeves), and Is Anyone
Responsible: How Television Frames Political Issues.

Don Green

“Gauging the Influence of
Political Advertising on Television and Radio: Results from a Large
Scale Randomized Experiment”

Abstract: We evaluate the effects of a $2 million dollar television and
radio advertising campaign on voters’ candidate preferences.
Television and radio ads on behalf of an incumbent gubernatorial
candidate were rolled out on a random basis in twenty media markets and
several dozen cable television markets over a three week period. Voter
preferences were measured using daily tracking polls (N=27,283) that
began prior to the start of the advertising campaign and finished after
its conclusion. Our results suggest that television
advertising had a powerful but short-lived effect on vote
preference. Radio’s effects are found to be weak and
insignificant, but radio’s effectiveness is not measured as precisely
as television’s effectiveness. We find no evidence that ads
are more effective in markets where the opponent fails to run competing
ads.

Bio: Donald P. Green is A. Whitney Griswold Professor of
Political Science at Yale University, where he has
taught since 1989. He also heads Yale’s Institution
for Social and Policy Studies, an interdisciplinary research
center that emphasizes field experimentation. His research
interests span a wide array of topics: voting behavior,
partisanship, campaign finance, research methodology, and hate
crime. His recent books include Partisan Hearts and
Minds: Political Parties and the Social Identities of Voters
(Yale University Press 2002) and Get Out the Vote!: How to
Increase Voter Turnout (Brookings Institution Press 2004).

Robert N.
Parker

“GIS and Social Research: Data Dissemination, Database
Construction, and Multivariate Analysis”
Abstract: Geographic Information Systems or GIS has the potential to
transform and enhance Social Science Research in a number of ways.
First, GIS can substantially enhance data dissemination in the
classroom, facilitate the presentation of scientific data to policy
makers, government officials, and the general public, and can be used
to enhance the presentation of data to scientific audiences. The
capabilities of GIS software created to support the construction and
manipulation of maps can also facilitate the construction of more
complex and potentially scientifically more useful data bases by
allowing researchers to combine geospatial information, geographic
structures, and other properties of space with what are traditionally
seen as non spatial data: behavior, attitudes, and characteristics of
geopolitical units. By creating the capability to merge these two
disparate types of data, GIS allows social science researchers to
discover new relationships, examine more interesting hypotheses, and
construct more effective explanatory models. Finally, geospatial
statistical models can address these more complex relationships,
hypotheses, and models; such models have been developed in spatial
econometrics and are necessary to properly accommodate spatial data and
structures.

Bio: Robert Nash Parker is Professor of Sociology and
Co-Director of the Presley Center for Crime and Justice Studies at the
University of California,Riverside. His research concerns the role of
alcohol, drugs, and substance abuse in violence, the causes of homicide
and youth homicide, and the design, implementation, and evaluation of
interventions which prevent or reduce youth violence and substance use.
His research has appeared in a wide variety of journals including the
Journal of the American Medical Association, The Journal of
Psychoactive Drugs, and New Directions in Evaluation, as well as the
usual sociological and criminological research journals. He is the
author of Alcohol and Homicide: A Deadly Combination of Two American
Traditions, the Co-Editor of a volume entitled “Pitfalls and Pratfalls:
Issues of null and negative findings in evaluating interventions,” and
the author of two forthcoming books, one on GIS methods in the Social
Sciences and the other on the Alcohol and Violence Relationship.

Sean
Reardon

“Propensity Score Matching for Causal Inference:
Possibilities, Limitations, and an Example”

Abstract:
Propensity
score matching has become increasingly popular in recent years as an
analytic
strategy for estimating the causal impact of a treatment in the absence
of
random assignment. In this talk, I will describe the
conceptual logic and
assumptions underlying propensity score matching (and other matching
strategies
as well). I will describe a body of research work that
attempts to
evaluate the bias in propensity score matching estimators. As
an example,
I will describe propensity score estimates of the
effect of
Catholic schooling (versus public schooling) in elementary school
achievement.

Bio: Sean Reardon is associate professor of education at
Stanford University, specializing
in the effects of educational policy on educational and social
inequality, the causes, patterns, and consequences of residential and
school segregation, and applied statistical methods for educational
research. His primary research examines the relative contribution of
family, school, and neighborhood environments to racial/ethnic and
socioeconomic achievement disparities. In addition, his research
develops applied quantitative methods for examining variation in
treatment effects and for measuring aspects of school and neighborhood
context. He teaches graduate courses in applied statistical methods,
with a particular emphasis on the application of experimental and
quasi-experimental methods to the investigation of issues of
educational policy and practice. Sean received his doctorate in
education in 1997 from Harvard University. He is currently a recipient
of a William T. Grant Foundation Scholar Award to fund his work on the
causal effect of neighborhood conditions on adolescent educational and
social outcomes. In addition, he is also a Carnegie Scholar, which
funds his work on the effects of programs for English language learners
on the educational trajectories of Latino students. Sean is also an
associate professor of sociology (by courtesy) at Stanford.

Jay McClelland

“Dynamical Models of Decision Making: Optimality, human performance, and principles of neural information processing’”

Abstract:
I will
present a model of the dynamics of decision making. The model can be
related to the theory of optimal decision making under conditions in
which noisy information is sampled continuously from the environment,
to details of human behavior seen in experiments requiring perceptual
classification or preferential choice, and to principles of neural
information processing. The process of model development and the
application of the model to experimental findings will be discussed,
and contrasts with other models attempting to address some of the same
phenomena will be presented.

Bio:
James L.
(Jay) McClelland received his Ph.D. in Cognitive Psychology from the
University of Pennsylvania in 1975. He served on the faculty of the
University of California, San Diego, before moving to Carnegie Mellon
in 1984, where he was a founding Co-Director of the Center for the
Neural Basis of Cognition, a joint project of Carnegie Mellon and the
University of Pittsburgh. In 2006 he moved to Stanford University,
where he is now Professor of Psychology and the founding Director of
the Center for Mind, Brain and Computation.
Over his career, McClelland has contributed to both the experimental
and theoretical literatures, most notably in the application of
connectionist/parallel distributed processing models to problems in
perception, cognitive development, language learning, and the
neurobiology of memory. He was a co-founder with David E. Rumelhart of
the Parallel Distributed Processing research group, and together with
Rumelhart he led the effort leading to the publication in 1986 of the
two-volume book, Parallel Distributed Processing, in which the parallel
distributed processing framework was laid out and applied to a wide
range of topics in cognitive psychology and cognitive neuroscience.
McClelland and Rumelhart jointly received the 1996 Distinguished
Scientific Contribution Award from the American Psychological
Association and several other awards for their pioneering work in this
area. McClelland is a member of the National Academy of Sciences, and
he has received the APS William James Fellow Award for lifetime
contributions to the basic science of psychology.

McClelland currently teaches cognitive neuroscience and conducts
research on learning, memory, conceptual development, spoken language,
decision making, and semantic cognition.

James Moody

“Diffusion over Dynamic Networks‘”

Abstract:
We often care about networks for the “bits” — information, money, disease,
support — that travel over the network. In this presentation, I review the
network foundations for diffusion and then extend classic treatments to dynamic
networks. The talk will cover a wide range of substantive topics in network
methods ranging from the local distributions of contacts to the global patterns
of connectivity evolving from such contacts.

Bio:

James Moody is associate
professor of sociology at Duke University. He has published broadly on the
dynamics of social networks, network methods, & theory. His work covers a
wide range of empirical settings from the dynamics of adolescent friendship
structure to the diffusion of sexually transmitted diseases. He is currently
working on questions related to the evolution of science networks, the structure
and dynamics of friendship networks, the foundations of belief systems and
network diffusion. He can be reached at jmoody77@soc.duke.edu.

Mollyann Brodie

“Using Public Opinion Research to inform Public Policy Debates in ‘Real Time’”

Abstract:
Researchers often plan and
implement studies aimed at informing public policy debates. However, ensuring that essential, in-depth
research findings are available in a policy relevant time frame produces
challenges to researchers in terms of the data collection, the analysis, and
the communication and dissemination strategies employed. This seminar will discuss how to effectively
produce and communicate in-depth surveys of public opinion and experiences to a
media and policy oriented audience in the course of ongoing highly visible
policy debates. We will use the recent
enactment and implementation of Medicare Part D – the Medicare prescription drug
benefit - as a case study.

Bio:
Mollyann Brodie is Vice President, Director of Public
Opinion and Media Research of the Henry J. Kaiser Family Foundation. Currently, she directs the Foundations’
survey research group as well as a variety of public knowledge and survey
related projects including ongoing survey partnerships with the Washington Post
and USA Today. Prior to joining the
Foundation, Dr. Brodie was a Health Policy Fellow and the Assistant Director of
the Program on Public Opinion and Health at the Harvard School of Public Health. Dr. Brodie currently serves on the Board of
Directors for the

Roper
Center

for Public Opinion
Research. She also serves as chair of
the Education Committee for the American Association of Public Opinion Research
and previously served on its Executive Council in 2004-2005 and as president of
its Pacific chapter in 2002. She
received her Ph.D. in Health Policy from

Harvard

University

.

Christian Wheeler

“Manipulation and Measurement of Construct Accessibility”

Abstract: Accessible constructs (e.g., traits,
stereotypes, goals, etc.) affect both judgments (e.g., impressions of
others) and behaviors. In this talk, I will discuss laboratory methods
used to investigate the influence of construct accessibility on
judgment and behavior. The talk will include examples of various
means of manipulating and assessing construct accessibility without
awareness on the part of the participant. I will provide
empirical examples of these techniques and discuss issues associated
with their implementation.

Bio: Christian Wheeler is
an Associate Professor of Marketing at Stanford University where he
teaches courses on Marketing Management, Attitudes and Persuasion, and
Research Methodology. He received his BA from the University of
Northern Iowa before moving to Ohio State, where he completed his MA
and PhD. His research has been published in top marketing,
organizational behavior, and psychology journals, including Journal of
Consumer Research, Journal of Personality and Social Psychology,
Organizational Behavior and Human Decision Processes, Personality and
Social Psychology Review, and Psychological Bulletin.

Paul Sniderman and Michael Tomz

“The Microfoundations of Issue Voting”

Abstract:
This paper, by Tomz and van Houweling, systematically examines how the
issue positions of candidates affect voting behavior. Our analysis
focuses on the three leading theories of issue voting: proximity,
discounting, and directional theories. We frst explain why existing
data and methods are insufficient to estimate the prevalence of these
three voting rules. We then formally derive, for the first time, an
exhaustive set of critical tests— situations in which the theories
predict different vote choices. Finally, through survey experiments, we
administer the tests to a nationally representative sample of adults.
We find that around 54% of voters apply a pure proximity decision rule.
Another significant proportion, 21%, discount the announced positions
of candidates by taking into account the location of the status quo.
Finally, about 10% of voters make choices consistent with a directional
decision rule, and the remaining 15% display behavior that is not
consistent with any of the three theories. These findings establish,
and at the same time qualify, the foundational assumptions in models of
democratic politics.

Bio:
Michael Tomz is Assistant Professor of Political Science. He
holds an M.Phil. from the University of Oxford, where he was a Marshall
Scholar, and a Ph.D. from Harvard University. Michael’s interests
include international relations, political economy, public opinion, and
statistical methods. His current research on the credibility of
international commitments is supported by a five-year CAREER grant from
the National Science Foundation. Michael is also engaged in
NSF-funded research with Paul Sniderman and Robert Van Houweling about
spatial reasoning in politics.

Paul Sniderman is Fairleigh S. Dickinson Jr., Professor in Public
Policy. He pioneered the introduction of randomized experiments
in public opinion surveys via computer-assisted interviewing. He
is P.I. of a NSF-funded research project on the behavioral foundations
of rational choice in politics; his Co-PI’s are Michael Tomz and
Rob van Houweling (of Berkeley).

Linda Piekarski

“Surveys in the 21st Century”

Abstract:
The quality of survey data has always been based on the quality of the
sample design and sampling frame(s), the quality of the questionnaire,
survey protocols and the response characteristics of the survey.
Technology is rapidly changing the way in which surveys are conducted
and will be conducted in the not too distant future. These changes are
impacting all of the quality measurements mentioned above.
This presentation will address a few of the myriad of issues associated
with building and maintaining different and overlapping frames, in
particular telephone frames, address frames and internet panels. It
will also discuss issues unique to each of these modes. Specifically, I
will address the challenges and constraints presented to each mode by
state and federal legislation, questionnaire design, declining response
rates and differential nonresponse.

Bio:
Linda
Piekarski has been with SSI since 1980. Linda currently
manages the Database and Telephone Production Groups. These departments
are responsible for creating, cleaning, and maintaining SSI’s numerous
telephone and Internet databases and for producing all telephone
samples. She holds a BA from
Wells

College
and an MA from

Yale
University

.

Linda also plays a key role in providing sample design
support both internally and to SSI’s clients. She has authored
numerous papers on sampling topics presented at industry conferences (AAPOR,
APDU, ASA, and CMOR) or published in industry journals. In recent years,
Linda’s investigations and presentations have focused on the challenges facing
the research industry such as declining response rates, the increased use of
cellular phones, and privacy concerns. She serves on the Technical Programs
Advisory Board of Gibbs College in

Norwalk
,
Connecticut

.