Andrew Kohut on Polling at the Pew Research Center

Andrew Kohut

President

Pew Research Center

Our current featured expert will be Andrew Kohut, president of the Pew Research Center in Washington, DC. Mr. Kohut also acts as director of the Pew Research Center for the People & the Press and the Pew Global Attitudes Project.

Questions and Answers

Question

How do you decide which matters to poll on? It’s easy to understand conducting surveys on the Iraq war or the presidential race, but how do you choose among more discretionary topics such as views of marriage or working mothers?

Answer

We are careful readers of the news, of course, and rely on our long experience in identifying the public opinion issues. We also turn to a kitchen cabinet of outside advisers, drawn from the media, polling and academia, who offer valuable input. We do our best to be far-sighted and to anticipate emerging issues.

Question

The Global Attitudes Project, which provides an international perspective through a series of worldwide public opinion surveys, must present different challenges than polling in the U.S. How would you characterize the differences: Is it more difficult to reach subjects? Is it difficult to survey enough people?

Answer

There are particular challenges to international surveys, but the principles of doing quality research are much the same—good reliable questioning, representative sampling, high-quality interviewing and thorough analysis of the data.

One of the keys is working with well-qualified local firms to conduct the fieldwork.  Our local partners often use a number of local languages in one country. And they employ the appropriate research methods to ensure that we get as representative a sample as possible. We conduct face-to-face interviews in most nations, since many parts of the world still lack universal access to telephones.

Question

The Pew Research Center completed the first-ever, nationwide survey of Muslim Americans, conducting more than 55,000 screening interviews to yield a valid sample of Muslims. What were some of the challenges you faced in undertaking this project?

Answer

The methodological approach employed was one of the most comprehensive and rigorous in the Center’s history. Interviews were conducted with 1,050 Muslims in English, Arabic, Urdu and Farsi. The random sample for the survey was built from three sampling sources. In total 57,549 households were screened to complete the sample.

Question

You’ve been polling for more than 30 years. How did you get your start? How has the profession changed since you began?

Answer

I was a graduate student in sociology at Rutgers, but I gave up my assistantship to work part time at Gallup. In the years since, the world of polling has grown up. It’s more sophisticated and more thorough. There’s more polling, in general, and unfortunately, there are also more examples of bad polling.

Question

Often polls validate findings that people intuitively know, such as the growing unpopularity of America’s involvement in Iraq. What are two or three examples of poll results that truly surprised you?

Answer

I was surprised by how the public repeatedly rallied to President Clinton—when the Lewinsky story first broke, and then again even when he admitted his wrongdoing. It showed the public’s ability to ignore the media, generally, and the pundits, specifically.

I was also surprised by the extent of bipartisan consensus that arose following the 9/11 attacks and then by how quickly it dissipated, leading to one of the most polarized eras in modern times.

Question

Your polls need to be timely, often responding to breaking news stories like Hurricane Katrina. How do you turn on a dime and publish the results so quickly while ensuring accuracy?

Answer

An important part of our job is getting in-depth public reactions to major events. We are always prepared to tear up our research plans to find out what Americans think when something big happens. For example, Pew Research was in the field by Sept. 13 to gauge the impact of the 9/11 attacks on the American public’s psyche.

Question

It’s well known that the precise phrasing of a survey question can radically change the answer. What do your pollsters do to ensure that questions are structured so as to avoid such distortions?

Answer

We rigorously test questions to find out if the questions are understandable and balanced. We routinely employ multiple versions of the same question when we are not quite sure about phrasing.  We also rely a lot on our battery of time-tested questions and formats for asking questions.

Question

An increasing number of young people only use cell phones, yet public opinion polling is still largely conducted through land lines. Do you worry about the accuracy of your results? Are you contemplating any changes in methodology?

Answer

We have been paying careful attention to the growing number of people who only have cell phones. Our studies show that currently the absence of this group from traditional telephone surveys has only a minimal impact on the results. But we realize that, in the long run, some effort will have to be made to include “the cell-onlys” in our polls. This year we have conducted a number of surveys with supplemental cell phone samples. We may have to do this more often and/or use other methods, such as mail and internet polling, to get to this slice of the population.

Question

What is your greatest hope and ambition for the Pew Research Center? What is your greatest challenge, i.e. what keeps you up at night?

Answer

Our greatest hope is to do a good job of informing the American public about the opinions and trends that are shaping their lives and the world—from politics to the media, to foreign policy, to the major social issues of our time. Our greatest challenge is to tell the story of public opinion fully and with nuance, and at the same time communicate it in a way that is understandable to our audiences.

Question

What are you most proud of in your lifetime achievements in polling and public opinion research?

Answer

I am proud of the leadership that I have provided to organizations that have been at the forefront of opinion polling in an era in which polling has come to play such an important role in American society. I was president of the Gallup Organization, founded Princeton Survey Research Associates, and serve as the first president of the Pew Research Center.