The people whose names are famous as pollsters — like Gallup, Harris, and Roper — can't and don't do the actual interviewing. For that job, they rely on hundreds of other less famous individuals who have a variety of backgrounds, qualifications and training.
Interviewers are the core of the survey process; they call you, talk with you and record your answers. Bad interviewers distort the way questions are asked, mispronounce the names of candidates, and inject their own opinions into the results. Good interviewers know what they are asking, are uniformly polite and stick to the interviewing script, so every person is interviewed the same way.
When modern polling began in the 1930s, interviewers went door to door in selected locations: interestingly, most were women. In a 1946 study, interviewers themselves were interviewed and asked what they looked for in a job. They wanted flexible hours, a job they could fit in to their daily lives, the freedom to accept or reject assignments — and extra income. In short, what they wanted was a part-time job, which is exactly what interviewing then provided.
Telephone interviewing supplanted door-to-door interviewing in the 1970s. At that point, more than 90 percent of all households could be reached by phone — and, coincidentally, rising crime rates and insecurity made door-to-door interviewing in urban areas almost impossible. Telephone interviewing also cost less and came with better quality control. Having interviewers in one central location means they all receive the same training. Phone calls are monitored to be sure that everyone is asked the same questions the same way. It's true that a recorded voice — or a Web-based questionnaire — would ask the same question the same way, but neither of those techniques sense when a respondent is having a difficult time with a question and can report that back to the poll manager.
That's where live interviewers really make a difference. They decide who should be interviewed, they have direct contact with the people interviewed, and they know when a poll question simply doesn't work — when respondents don't understand it. Interviewing today can be a full-time job, and the proportion of male interviewers has increased: among the active interviewers who now work on CBS News Polls, slightly more than half are men.
There has been a wealth of research on how to ask questions, and on the impact individual interviewers can have on the process. Public Opinion Quarterly, the official journal of the American Association for Public Opinion Research, has published more than 50 articles about interviewers, how to train them, and their impact on respondents. There have been studies of interviewers' voices, their verbal idiosyncrasies, their attitudes and their demographic characteristics. Do women interviewers get different answers on questions about women's rights than male interviewers do? What kinds of respondents — men or women — give different answers to male and female interviewers? In this election, with a woman, Hillary Clinton, currently the front-running Democratic candidate for president, analysis of questions like these should become routine.
Interviewers for the CBS News/New York Times Poll work at our own interviewing facility in New York. They make their calls from a room in the CBS News Broadcast Center that has been equipped with more than 50 phone lines and computers, just a few floors away from the poll's offices. This ensures frequent communication and feedback between interviewers and analysts; it also provides the flexibility to react quickly to news events.
Our interviewers are among the most experienced in the business. They reflect the racial and ethnic composition of the New York area. Some are retired from other jobs, some work as interviewers full-time, both for us and for other interviewing facilities, and some are actors who use interviewing as a way of financing the down times in their careers. Nearly half of our current active interviewers have worked for CBS News and the Times for more than eight years. Our interviewers are a dedicated and talented group.
I'm thinking a lot about interviewers this week, because on Sunday I attended the funeral of one of CBS News' long-time interviewers and supervisors: Michael Bennett, who passed away this month. He worked with us for about 20 years, a time that encompassed the change from writing answers and circulating random phone numbers on paper to using computer-assisted techniques. He interviewed respondents. He took verbatim answers and comments that respondents gave and classified them by topic. For more than 15 years, he was our main supervisor during daytime interviewing hours. It was he who handled calls from respondents who needed reassurance that we really were who we said we were.
Michael Bennett may not be a famous name in polling. But he was one of the unsung people who really matter in our profession, and we will miss him.
By Kathy Frankovic