Sometimes things are too good (or just too strange) to be true. Like “fake news.” Nearly 90 million pages in a Google search reference it, and there are plenty of instructions about how to discover when news is fake (this is from NPR; this from the University of Michigan). But what about “fake polls?”
Fake polls exist and they have been around for years. We have reports that they have become part of elections in many countries, including Mexico, and as early as 2000. “Fake polls” have their own memes.
Sometimes fake polls are completely manufactured, like this one that said Kid Rock would beat a sitting U.S. Senator. But sometimes they seem to produce real results, labelled as having been done by a known research company.
They are a global concern. And they are of special concern to the research industry. Fake polls attempt to mislead both readers and respondents and undercut the confidence that good polls need. They may cause people not to answer real polls. And they undermine democracy. Organizations like ESOMAR, WAPOR and AAPOR care about ethical standards and set the rules that real researchers are required to follow.
How can you tell something is a fake poll? Here are some guidelines:
For example, if mail arrives that asks you to indicate the national issue you care most about, and then asks you for a contribution to a political party or interest group or charity, THAT IS NOT A REAL POLL.
If you get a phone call that asks you about your health and then suggests that you purchase a device that is supposed to help you with your problem, THAT IS NOT A REAL POLL.
If you see a website with a single “pop-up” question, THAT IS NOT A REAL POLL.
If you read a result that appears too conveniently, at a time too quickly after an event, with no indication of how it was conducted, THAT IS ALSO PROBABLY NOT A REAL POLL.
A fake poll can masquerade as a real poll by claiming it was conducted be an actual poll company, when there was no such poll. Beware of polls found only on social media, where results are distributed through contact networks without any control, usually with no information about how exactly it was done.
Fundraisers, campaigners and unscrupulous marketers can use fake polls to attract attention.
Recently, President Trump’s former personal lawyer testified to Congress that he paid to have internet “polls” rigged to favor Donald Trump as a presidential candidate. The internet “polls” weren’t really polls at all, just website votes. The person Michael Cohen, the lawyer, hired wasn’t even a pollster. What Michael Cohen paid this person (whom The Wall Street Journal described as running “a small technology company) to do was write a script to cast multiple votes in a website tally. THAT, OF COURSE, WASN’T A REAL POLL.
“Fake news” and “fake polls” aren’t just fakes. The terms have taken on a politically charged meaning: politicians use the term to describe any news or any polls they don’t like. U.S. President Donald Trump talked about “fake news” in 318 tweets in the first two years of his presidency; he also extends his attack to real polls. If someone doesn’t like a poll result, they just call it “fake.”
The existence of fake polls is one thing – the claim that anything a politician doesn’t like is a fake poll is quite another and makes polling all that much harder.
There are two important behaviors that both researchers and the public need to understand: how to avoid becoming part of a fake/scam poll and how to recognize a report of a poll that is fake.
1. HOW TO AVOID BECOMING A FAKE/SCAM POLL STATISTIC:
A big problem is that a lot of things that aren’t polls are called polls. ESOMAR (along with other organizations like WAPOR and AAPOR) has condemned practices like “sugging” (selling under the guise of research) and “frugging” (fund-raising under the guise of research). These truly fake polls use a survey only as an excuse for collecting money. In the United States even the most virtuous organizations have been known to do this, believing this is the best way to get people interested in what you have to say (Researchers should probably accept this as a backhanded endorsement for real public opinion polls).
“Sugging” and “frugging” attempts reach their victims by mail, by telephone or on line. They ask relatively few questions but they always ask for money: to donate to a charity or to a political candidate, or to make a purchase. Sometimes they ask for personal data like identification number, or bank account and credit card information. THESE ARE NOT POLLS! In fact, they are scams.
No reputable polling organization would conduct these scam polls. They are condemned by ESOMAR, WAPOR and AAPOR.
There is also campaigning under the guise of research, using fake polls for fundraising appeals (“frugging”) or by making false statements. This type of fake poll contacts many people, asking only a few question, using ns, each containing negative information about an opponent. These are known as “push polls,” which aren’t really polls at all. As CBS News wrote, “A push poll is political telemarketing masquerading as a poll. No one is really collecting information. No one will analyze the data. You can tell a push poll because it is very short, even too short. (It has to be very short to reach tens of thousands of potential voters, one by one). It will not include any demographic questions. The “interviewer” will sometimes ask to speak to a specific voter by name. And, of course, a push poll will contain negative information – sometimes truthful, sometimes not – about the opponent.”
You can consider push polls as political advertising to individual voters. Today, those “fake polls” can be taken on Facebook, through Twitter, or by other social media. They pass on false information, reaching voters one by one, exposing them to a campaign message that may or may not be true. A real poll may sometimes test negative message about a candidate, but a real poll will ask much more.
These very short fake polls are much like the internet trolls who have interfered in elections in the United States, in the United Kingdom and elsewhere, posting pictures, “news,” and polls, on social media making arguments that may or (more likely) may not be true.
What can you do?
If you think something is not right, it probably isn’t. Transparency is mandatory when collecting personal data. If it doesn’t exist, DON’T ANSWER.
Don’t answer to anything that claims to be a poll or survey if any of these conditions apply:
- You are asked for cash or some other sort of action.
- The company is unknown or the interviewer’s behavior seems unprofessional.
- There is no promise of confidentiality or the company refuses to provide information about its rules for data privacy. Researchers must have a privacy notice that is readily accessible by data subjects and is easily understood.
These are the polling scams – tricking people into thinking they are answering a real poll where researchers care about their opinions when they are not.
2. HOW CAN YOU DETERMINE WHETHER A REPORTED POLL IS REAL OR FAKE?
There are a few tests one can apply – and frankly, one should apply every time you read or hear about a poll finding. This is not a long checklist, but a set of rules that are basically common sense.
- Consider the source. Who is reporting the poll? Have you heard of the survey organization? Is it coming from a known (and reliable) news source? Does the person reporting the poll – or paying for it — have a political bias? Real polls should appear on an organization’s website. Can you tell if the pollster is a member of international organizations like ESOMAR and WAPOR?
- Read the story. Does it make sense? Could the poll actually have been conducted in the time period described? Are you told what group was part of the sample (are the opinion supposed to be of all adults, of people who voted, of people who tweeted, or people who gave money to a candidate)? Was the question or questions they asked reasonable? Reputable polling organizations are required to report their methodology, and real journalists should be asking the same kind of questions. There is a free course for journalists on how to report on polls, developed by AAPOR, ESOMAR and WAPOR. Does something strike you as wrong? If a figure is really surprising, you might want to look at other polls to see if their results are similar.
- Ask questions. Real pollsters will answer them. In fact, they must answer them, according to the ethics guidelines of associations like ESOMAR and WAPOR.
In short: Consider, Read, Ask. And you should report any polls you believe are fake to your local association, ESOMAR or WAPOR.
