AAPOR
The leading association
of public opinion and
survey research professionals
American Association for Public Opinion Research

2007 Presidential Address

2007 Presidential Address from the AAPOR 62nd Annual Conference

Gladly would We Learn and Teach – and Gladly should We Gather and Preach by Robert P. Daves

I am humbled to stand before this wonderful group of colleagues, teachers, associates – and yes, many, many friends – and carry out an AAPOR tradition of more than a half century. In 1950, the year I was born and four years after the first AAPOR gathering in Central City, CO, our founders instituted a “Presidential Evening” as an occasion on which the Association takes stock of some of its major problems. This occasion has become a custom where members grant the current president a few minutes of the conference's time to share a few of his or her thoughts.

Let me share something that I have learned about us in the past year. One is that despite having many gardeners in the ranks, there are no shrinking violets: You have not been shy in sharing your opinions with me. And I appreciate that. That trait apparently has been passed down through generations, as past presidents have been quite ready in their remarks to suggest what future presidents should do, as well as that “stock taking.”

But before I do that, let us see what a few recent presidents have suggested and what AAPOR has done to implement their suggestions.

Mike Kagay in 1999 documented a disconnect between elites and public opinion (Kagay 1999). I believe that public opinion polls continue to document that disparity, allowing voters in the 2006 elections to know that many others felt as they did – dissatisfied with the course of the country. We have heard several times at this conference that the disconnect is worldwide, and that it continues in the United States. (Pew Research Center's Pew Global Attitudes Project, Spring 2007). One needs to look only at how voters reacted at the polls in 2006, both at the statehouse and national levels, to see that their opinion – which we as pollsters have documented over the past number of years – was at odds with the country's leadership. They threw many of those leaders out.

We have made some strides in educating journalists, as Don Dillman(2002), Mike Traugott (2000), and Cliff Zukin (2006) suggested. AAPOR now has a curriculum for a short course for journalists on polling that has been pilot tested and delivered to more than a half-dozen groups over the past several years. But that is only a drop in the bucket. We will extend our educational reach by a new relationship with the Poynter Institute for Media Studies to help Poynter train journalists when they cover polls in the 2008 election.

We have not done too well for one of Diane Colasanto's recommendations – tracking the use of polling data in public policy debate (Colasanto 1997). But one of her other recommendations, speaking forcefully about AAPOR's standards through press releases, is something that we have been able to accomplish more and more during the past several years – witness AAPOR's reaction through Cliff Zukin's efforts last year to the American Medical Association's bogus spring break poll. Then there was Nancy Mathiowetz's response recently to the less-than-rigorous MSNBC poll on workplace attitudes.

It is hard to determine how well researchers have done on Mark Schulman's charge to be seen more as problem solvers rather than technicians (Schulman 2003). Personally, I have taken it to heart when I worked with internal newspaper and website clients. Both Mark and Betsy Martin cautioned about the tyranny of numbers and the potential dangers of going to extremes to boost response rates. And since Betsy made her plea for more work on nonresponse effect, we have seen a special edition of Public Opinion Quarterly devoted especially to nonresponse (Martin 2004).

Now that we have looked at what a few past presidents have suggested – and how well we have done meeting their calls – I would like to focus on two specific challenges I see for us as AAPOR members. One of the challenges is external, looking outward, and the other is internal, looking at our beloved organization.

Let us take a look at AAPOR itself first, and “take some stock” as the founding members suggested. In the past few years, AAPOR has grown. Please join me in applauding ourselves, because this year for the first time our membership passed the 2,000 mark. Also, AAPOR's budget, much of which deals with the conference, is nearly a million dollars.

With that growth come additional resources, and I believe Council has allocated those resources wisely. Some of its initiatives include a new Internet magazine called Survey Practice, scheduled for launch this fall. There also have been special editions of POQ. One powerful initiative is an enhanced communications capacity – including a professionally redesigned web site.

In addition to these communications initiatives, Council has enhanced insurance protection for chapters and their leaders, something that needed doing in these litigious times. Council also provided the resources to help them incorporate if they wished, and to seek nonprofit status. That is a lot, and all of these things are crucial for achieving AAPOR's long-range strategic goals as an organization. But there is one value that we always acknowledge, always lament as being lost, but then not much is done about it.

Many of us will remember the little customs that have fallen by the wayside over the years. Singing “Mares Eat Oats” and other old songs at the Saturday night sing-along. Going to smaller conference sites at out of the way places where there were no big-city lures or flashy distractions to keep us from learning and teaching each other about this marvelous thing called public opinion research.

Many of those are lost. This year, we have about 800 researchers registered at the conference. Professional meeting planners tell us that there are only a limited number of hotels that will work for conferences “our size.” We feel obligated – correctly so, in my opinion – to visit different parts of the country, which may further limit some of our site choices.

In many ways, we need to have a larger conference. Much more work is being done now in survey methodology, measurement of topical issues, and the problems we all face such as mode effect and coverage issues because of cell phones. We need to hear about that research, and want to hear about it before it is published, so we can incorporate those teachings into our work. That suggests an expanded, bigger conference.

Our conference planners have done a great job with this conference. They could have partitioned the cell-phone track into separate days apart from the conference. But they decided that it would be better to fold the track into the conference so researchers mainly interested in cell-phone research also could benefit from the wider conference. That was the right thing to do, but we must do more.

Murray Edelman alluded to the sense of community that AAPOR fosters (Edelman 2001). During my term as president, I was greeted warmly any time I called a member on behalf of AAPOR because of the good will and respect we have for each other. There were very few times when I asked for something on behalf of AAPOR and was turned down, and then only because there were prior commitments. When I showed up at my first AAPOR conferences two decades ago, I was welcomed as a valued colleague by AAPOR stalwarts such as Glen Roberts and Harry O’Neill. Many others have been quick to help with questions about methods and measures.

If you have not figured it out by now, AAPOR means a lot to me. You are my professional community, and you have come through for me time and again. We have a wonderful organization, and we have good people in it. But let me suggest two things.

First, if you have not yet, get involved in AAPOR. By working with other researchers, you will get to know some wonderful folks, learn more about our art and science, and contribute. Our conference folks always need an extra hand, and that is certainly a good way to get involved. Just as importantly, we should continue to experiment with the structure of our conference, as we have this year, balancing the need to learn and teach with the need to foster the sense of collegiality, community, and commitment to our profession.

But what about the lions and tigers and bears that are on the prowl for public opinion research and our practitioners?

Over the past decade – perhaps less time than that – we have seen a sea change in the media landscape. The rise in broadband capability and penetration has endowed consumers with relatively inexpensive and reliable high-speed access to the Internet. Broadband access lets us surf the Internet on the big waves. We can swiftly switch from site to site to find out the daily squabble in the blogosphere. We can easily consume high-density content such as video and audio, and flash graphics. And Wi-Fi has given us the flexibility to do all of this from a wide variety of places. In my little chunk of heaven in flyover land called Minnesota, several cities and towns are nearly ready to implement cheap Wi-Fi access citywide, available from park bench to private back porch. In other words, in many of the important places where public discourse takes place.

Surfing is easier and faster, and search engines are more clever and powerful, making it easy for users to find quick answers to their questions. Much good has come of this. For example, think how easy it is to search the back copies of POQ in JSTOR. But as with any tool, there are those who use the Internet irresponsibly for mean-spirited mischief.

Most of you know that former AAPOR president Jim Beninger is the father of AAPORnet, one of the wonderful benefits that tie us together. When he had this podium in 1998, Internet penetration was less than 40 percent; now it is more than 70 percent, the good folks at the Pew Center tell us (Madden 2006). Jim might not have heard the word “blogosphere” back then, but he certainly would have understood it. The Internet has affected virtually all traditional news organizations, including national broadcast and cable news organizations, and national and large regional newspapers, in several dramatic ways.

One is a serious loss of advertising revenue. News organizations are a business, and many are in the process of sorting through what they can and cannot afford any more. Some already have made the decision, cutting staffs, and trimming – somewhat sharply – nonlabor budgets such as promotion and market research – and polling. I am a victim of those cuts. Last week, the Star Tribune, the nation's 10th largest newspaper, told me that it was eliminating my job as the director of the Minnesota Poll, a position and poll that the newspaper had funded since 1944.

Faster, cheaper, more widespread Internet access also has had a major change on the way we get our news – and what we do about it. Just think. A decade ago – perhaps less than that – most of us would get our news from either the newspapers or the broadcast network news. What would we do with that information? Mass media uses and gratifications research that dates back more than three decades tells us. We would consume news because it was intellectually interesting and entertaining. We would use it for water-cooler or back fence talk – that is “social facilitation” in the literature. We would use it to help make civic decisions. We would use it for consumer decisions, such as shopping (Blumler and Katz 1974).

But now those readers, viewers and listeners also are surfers. And more and more of them are bloggers – citizen journalists who wield their mighty virtual pens on the Internet. Conventional wisdom says there is a blog started every second, and the wags say that most of those have a readership of two – the blogger and sometimes the blogger's partner.

Many blogs, however, are nearing the point – both in terms of readership and impact on public discourse – that they have the power that the mainstream media has. One of the things that they are doing with their newfound power is attacking those with whom they take political issue. And that includes us, the pollsters.

Of course attacking pollsters is not new. One of the earliest recorded attacks on a poll was in the 1824 presidential campaign, when straw polls at taverns and militia musters were common (Littlewood 1998). One of the biggest complaints was about sampling – an attack familiar to us nowadays. Then the complaint was about the bias that comes from polling drunks at taverns. That is changed, with the bloggers and many political officials claiming personal bias and political agenda on the part of the social scientist who directed the poll.

Who among us has suffered these personal attacks?

More than 10 years ago, political leaders in New Jersey State excoriated the Star-Ledger Eagleton Poll and the pollster when her poll found something uncomplimentary about legislation it was considering.

About that same time a University of New Hampshire pollster was attacked for polling he had done in New England. I organized a panel at an AAPOR conference called “Pummeling the Pollsters” to begin documenting the problem.

In 2004, MoveOn.org took out a full-page ad in the New York Times, attacking George Gallup, Jr., the retired Gallup co-chair who had nothing to do with the poll. It was a vicious, mean-spirited assault launched because MoveOn.org claimed that the poll had too many partisans of one stripe in the sample. It did not.

In 2006, the polling director of the Los Angeles Times was singled out, with the attacker calling her poll a mess, and suggesting that she was incompetent. The poll was fine; she is incredibly competent. Shortly after her poll results were published, other national polls began finding exactly the same thing.

Also in 2004, some partisans' frustration with the Star Tribune boiled over after it published a poll that showed George W. Bush far behind John Kerry in the state. At noon one sunny September day, about three or four dozen protesters showed up at the Star Tribune’s front door – right after the TV cameras – carrying signs and shouting, “Hey, Hey! Ho, Ho! Rob Daves has got to go!” They had been organized by partisans devoted to tearing down any message or messenger deemed detrimental to its favored candidates. Both the Gallup Poll and the Minnesota Poll turned out to be not just close, but spot-on target, with some of the most accurate polling done in the election that year. And in fact, accuracy is part of the antidote to this criticism.

Of course it is not news, and usually not a very big deal when politicians pummel pollsters for publishing data they do not like. But those kinds of attacks are unconscionable, crossing the lines of propriety, good manners, civil discourse and ethics. In our 2005 POQ article Frank Newport and I recommended a number of steps that researchers can take to help. That includes hewing to – or exceeding – AAPOR's disclosure standards, better reporting on polls, responses by researchers and their organizations to the attacks, and support from professional organizations, such as AAPOR (Daves and Frank 2005).

Polling companies and news organizations ought to be able to take care of themselves. And they ought to have the public relations guts to do whatever is needed to protect their researchers. These companies have professional organizations to help them. CASRO, the Council for American Survey Research Organizations, can and does come to the defense of its members. But CASRO does not represent individuals. That is where AAPOR should come in.

AAPOR is an organization of individuals. When we sign the Code of Professional Ethics and Practices, we sign it as individuals, not as representatives of the Gallup Organization, or the Star Tribune, or the Los Angeles Times, Princeton Survey Research, or any other organization. We agree in writing to use only the most appropriate research techniques that are suited to the problem at hand, and to disclose our methods fully.

AAPOR has an increased communications capacity, and we are doing better at responding to bad surveys that make the news. We are doing better at getting AAPOR's standards of disclosure out in front of the public. And with the redesign of our website and our education programs and partnerships in the works, we will be doing better at helping journalists sort through some of the issues.

In the past few years on Council, I have watched membership chairs help us frame many of our decisions by asking, “How will it help our members?” We can take that tack here, too: “How can AAPOR help its members who suffer these attacks?”

Now that AAPOR can react more swiftly to the news of the day, I propose that we should not limit ourselves to serving as quick contacts for journalists with a question about a poll, or making quick responses about bad polls, as we do now. I would propose that we should feel obligated to come to the defense of any member who is singled out and attacked in the course of doing their research.

Yes, there are questions. Was the research, in fact, good work? That has to be asked, and there are ways of addressing the answer through a standards complaint. But standards investigations have their own problems, not the least of which is that they take a long time, and the moment to be most effective is long past when the investigation results are finally made public. Even then, there may be little help for the pollster who was attacked.

That does not mean we should abandon standards investigations. Far from it, we should continue to rigorously and swiftly investigate standards complaints. But we should do more.

AAPOR should stand behind its members when the situation warrants it. And members should speak out for each other when that is warranted. We could suggest publicly and swiftly that the researcher signed the agreement to use appropriate methods, and that any attack should be based in fact, rather than ill-informed, mean-spirited name-calling.

Several weeks ago we took the first step along this path when I sent out a press release condemning Ann Coulter's attack on a Newsweek Poll, and by extension an attack on the fine researchers at Princeton Survey Research Associates and Braun Research.

Any of us in AAPOR – any of us – could find ourselves the target of an ad hominem attack at any time. Media pollsters are the most visible, and it is not too big of an exaggeration to say that many walk around with a target on their backs. But as you know, many of us work for news organizations, market research companies, or public affairs consultants that also find our work in the public light. All it takes is a passionate, clever, and sufficiently unethical or ignorant person or group to begin pummeling the pollster, trying to deflect the attention from the findings to the researcher.

I challenge future Executive Councils – and all AAPOR members – to take up that task of helping defend each other when we are unjustly attacked.

If I read my AAPOR history right, the first presidential address came from Paul Lazarsfeld in 1950, and was published in POQ in 1951. He reviewed a number of challenges facing pollsters. Among those challenges were deciding topics and variables to include on questionnaires. He suggested including such newfangled techniques such as an index of political participation, interest in politics, and many other things that we do as a matter of course today.

But Dr. Lazarsfeld completed his speech by noting something that is still crucial for AAPOR today, and for its future. “We want all of our intelligent fellow citizens,” he said, “to have respect for the kind of work we are doing.” His advice? Show that we recognize our common problems and contribute to their clarification (Lazarsfeld 1951).

The sessions so far on cell phones have convinced me that clever and smart AAPOR researchers will help us figure out how to deal with the cell-phone coverage problem. We already have hints that nonresponse problems may not be as much of a bogeyman of bias that we feared. And we will continue to work on other sources of error in sample surveys.

The Council for Marketing and Opinion Research, CMOR, and CASRO are working in the legal and legislative vineyards to protect researchers and help respondents understand why cooperation on surveys is so important. But we must blunt the common problem of attacking our research and our researchers as a public relations tactic in a critic's or party's playbook. As individuals and as an organization, we can do that by sticking up for our members – publicly and privately – when they are under attack.

In many ways, we members of AAPOR are kindred to Geoffrey Chaucer's scholar, one of the characters in the author's Canterbury Tales. Gladly would he learn, and gladly teach, Chaucer described him. We could add to that and build a description of public opinion researchers: Gladly would we preach – preach the gospel of good research, good researchers, and convocations like this one where we can learn, teach and bind ourselves together. As Ben Franklin said, “We must hang together, lest we all hang separately.”

One of the ways we can hang together is to take the pride in this wonderful organization called AAPOR that we all feel, and let others know about it. Thank you very much for your attention today – and for allowing me to help lead the Great Ship AAPOR for the past year in such interesting and promising times.