Looking Ahead

As SIESR grows, we envision expanding the Institute's work in several different directions.

Documenting the accuracy of survey data

If a person were to ask, "How accurate are surveys?", there is no place to look for a solid answer. A search online yields unauthored blog posts asserting that well-designed surveys are generally accurate, without providing any evidence. This needs to be remedied. We will bring together a large body of evidence documenting both the accuracy of people's answers to questions and the accuracy of systematically drawn samples in describing the populations they come from, and give people a reliable answer to the question of why they should trust well-conducted surveys.

Documenting the widespread uses of surveys

Surveys are used for numerous important purposes and have tremendous impact in many arenas of life. But anyone interested in learning about those uses has no one-stop shop to visit. We will collect and present documentation of how surveys are used by governments, by NGOs, by businesses, and by academics — so that when people need information that survey data can provide, they can rely on it confidently, and when they are invited to participate in a survey, they can be confident their time is well spent.

Guidance for courts, juries, and journalists

A great deal of litigation involves adjudicating claims about large groups of people, and courts have come to realize that scientific surveys can substantially streamline the process of generating reliable evidence. But each time a survey is submitted to a court, experts on opposite sides argue about the quality of the methods, and judges and juries are asked to decide whether to rely on it. These decisions have often been made haphazardly: one court has ruled a survey's response rate unacceptably low when another court ruled a different survey with the same response rate acceptable. Judges and juries need guidance on how to evaluate surveys, and so do journalists working on deadline. We will provide it.

Informing voting decisions by legislators

Elected representatives pay attention to what their own constituents believe and want, but surveys of the opinions of people in a single state or Congressional district are as expensive as national surveys and rarely occur using optimal methods. Sophisticated statistical methods developed recently at Stanford allow us to concatenate the data from a large set of high-quality national surveys and disaggregate it into results for individual states. We will produce maps of public opinion in each state on a wide variety of issues.

A biannual conference on survey methodology

Conferences on cutting-edge methodological questions are regularly organized ad hoc and attract large numbers of attendees, but no regular schedule assures participants of a forum for coordination and insight sharing. We will host a biannual conference bringing together leaders in the field, and consolidate their presentations into a series of books published by a prestigious press.

A virtual faculty in survey methods

No single university has the world-class faculty that could be created if survey methods experts from around the country were coordinated into a single enterprise. By bringing faculty and practitioners to Stanford each year to deliver guest lectures, hold seminars, and conduct individual consultations, we can offer our students a better learning experience than is possible anywhere else.

A faculty fellows program

Modeled after the Robert Wood Johnson Scholars program, our vision is to recruit cohorts of scholars across disciplines, give them in-depth, top-shelf training in all aspects of survey methodology, and fund high-quality research by them — so that over the years these cohorts become an increasingly sizable body of scholars who do superb research using survey methods and provide superb training to others, helping to establish best practices as the norm in academia, government, and the private sector.

A team conducting original research

A full-time staff, together with post-docs, graduate students, pre-docs, and undergraduates, conducting cutting edge research on survey methodology for wide dissemination, tracking the methodological and substantive survey research being done around the world. These studies would be both experimental, involving collection of new data, and archival, involving analysis of previously collected data, seeking to understand how the method works in practice and how to optimize its performance.

Partnerships

An important feature of the Institute will be partnerships with organizations interested in survey research, to enrich the experiences of both faculty and students. Organizations that conduct, analyze, and report surveys continually throughout the year can provide settings for student training, collaborators in studies, and funding for joint ventures.

Supporting this work

There is no obvious funding mechanism to support a center like this one. Social science funding agencies support specifically designed research projects with substantive emphases, not methodologically focused centers for academic leadership with such broad and overarching goals. An institute like ours may well be able to support itself through consulting and charitable contributions in the long run. But for it to demonstrate its value, start-up support is required. Inquiries may be directed to iriss-info [at] stanford.edu.