We asked people

Surveys are another common source of evidence because they give the impression of a consensus about an issue.

For example, this article uses survey data as evidence for popular opinion about climate change:

People’s views about climate change, from how worried they are about it affecting them to how willing they are to do something about it, have shifted in developed countries around the world in recent years, a new survey by the Pew Research Center finds.

The study polled more than 16,000 adults in 17 countries considered to be advanced economies. Many of these countries have been large contributors to climate change and will be expected to lead the way in fixing it.

In general, the survey found that a majority of people are concerned about global climate change and are willing to make lifestyle changes to reduce its effects.

However, underneath this broad pattern lie more complicated trends, such as doubt that the international community can effectively reduce climate change and deep ideological divides that can hinder the transition to cleaner energy and a climate-friendly world. The survey also reveals an important disconnect between people’s attitudes and the enormity of the challenge climate change poses.

This article is actually a report, not an argument.

  • In that context, the claim is that, in developed countries, people's views about climate change have shifted in recent years.
  • The evidence is the findings from the survey.

Notice that this snippet doesn't give us any concrete data apart from a count of the respondents and their locations.

Rather, the findings are summarised (and we have to take it on trust that the survey was well-conducted and the conclusions accurately presented here).

This snippet uses survey data to argue that Americans associate science with hope:

ScienceCounts, a nonprofit organization working to strengthen public support for science  which I collaborate with, conducted a couple of polls that ask respondents a multiple choice question about what comes to mind when they hear the word “science.” What they found was clear: The U.S. public feels “hope.”

In the 2018 ScienceCounts survey, 63% of respondents said when they hear the word “science,” “hope” comes to mind. The next most common responses, at only 9% and 6%, were “fear” and “joy.”

More important, the feeling of “hope” held across different demographics, regardless of political ideology. A survey scheduled for fall 2020 will test if these associations still remain, amid the coronavirus pandemic.

The article even presents some of the data as a chart:

Notice how this snippet mentions actual data from the survey: 63% of respondents vs 9% of respondents, etc.

  • The issue or question is about what comes to mind when Americans hear the word "science".
  • The claim is that they feel "hope".
  • The evidence is attribution, summary, and data.
  • This snippet also limits the claim by suggesting that the results may not hold after the pandemic.

However, if surveys collect quotes and stories, people will often quote those as evidence (in the same way that we might quote experts or observations):

Our respondents told us they have strong aspirations for their futures. As one student told us:

"[I want] to grow my [business] into a popular […] local gardening landscaping service supporting houses and gardens that need repairing."

Others told us they want to become “totally independent” and talked about pursuing happiness and community connections:

"[I want to] find some work I purely enjoy and just live a normal life. Preferably doing some trade work or contributing to the community."

Young people also specifically spoke about doing further study and going to TAFE or university, to do a wide range of courses from computer science to education and medicine:

"I aspire to be a paramedic or doctor when the time comes, I would like to go to university to study medicine."

The snippet interleaves summaries with direct quotes from the survey.

Notice how the snippet doesn't attribute the quotes? That's because most surveys are anonymous.

This article, about the complex relationships between young people, adults, and online abuse, mixes summary and quote even more:

Young people are often reluctant to involve adults in their online lives. Many fear that parents and teachers will misunderstand or “overreact” in response to what they mostly regard as normal, unproblematic behaviour and experiences. Others say they are frustrated by adults who “trivialise” their experiences.

Over the past eight years, I have had extensive discussions with (mainly) teenagers from a diverse range of social and economic backgrounds, ethnicities, sexual orientations and genders about their experiences of social media and messaging apps. A lot of those I speak to initially try to downplay any issues. They make it clear they like being online and know how to handle any problems that may come up.

But when I ask them to tell me more about these problems – while remaining neutral and interested rather than appearing judgmental – it’s almost like the floodgates open. They want to talk about the things they don’t like and struggle with; they just worry that they’ll get into trouble if they are too honest.

Some describe a relentless stream of abuse and hate that can “ruin” the experience of being online. One 14-year-old girl says there is “so much sexism, racism, homophobia” which she thinks is wrong, but at the same time just an inevitable part of being online. A 14-year-old boy discloses: “Sometimes they’ve been racist to me … Racist comments [in] messages from other people.”

Some LGBTQ+ girls tell me about the extent of hate they experience online:

"[There’s] a lot of bullying … it’s coming from both adults and other children, [even] in safe spaces. There’s group chats online where people are added and it’s purposely [so people can] hate them."

This snippet is mostly a summary of the survey results, but it scatters fragmentary quotes throughout.

And notice how this snippet attributes some of the quotes, but it uses demographic features like age and gender rather than names. (Again, survey data is usually anonymous.)

Of course, like studies, surveys can be badly designed and the results can be misused.

One notorious example comes from mid-century cigarette advertising.

For almost a decade, Camels Cigarettes ran an ad campaign based on the claim that "More Doctors Smoke Camels":

More doctors smoke Camels Survey

"More Doctors Smoke Camels" is a factual claim. But is it true?

To substantiate the claim, the fine print of the ad uses a survey as evidence:

According to a Nationwide survey:


Like the rest of us, doctors smoke for pleasure. Their taste recognizes and appreciates full flavor and cool mildness just as yours does.

And when 113,597 doctors were asked to name the cigarette they smoked, more doctors named Camels than any other brand.

Three nationally known independent research organizations conducted the survey. They queried doctors in every branch of medicine.

But here's what Stanford University had to say about the method for getting this survey data:

In an attempt to substantiate the “More Doctors” claim, R.J. Reynolds paid for surveys to be conducted during medical conventions using two survey methods: Doctors were gifted free packs of Camel cigarettes at tobacco company booths and them upon exiting the exhibit hall, were then immediately asked to indicate their favorite brand or were asked which cigarette they carried in their pocket.

That means one person gave doctors packets of Camels, then shortly afterwards another person asked the doctors what cigarettes the doctors had—which, unsurprisingly, were Camels!

Let's take a leaf from Camel's book and make up some fake survey data.

Here's a headline:

Use made-up survey data as evidence to support this claim. Use at least 3 pieces of fake survey data, including one count, one percentage, and one quote.

Use made-up survey data as evidence to support this claim. Use at least 3 pieces of fake survey data, including one count, one percentage, and one quote.

Here's an example of some made-up survey data:

The annual "Southern Satisfaction Survey" gathered responses from 10,000 Australians across various states and revealed that only 35% of respondents rated their life as "bearable" or higher, a stark contrast to the 60% recorded in the previous decade.

Additionally, the number of individuals reporting high levels of daily stress jumped to 5,500, up from 3,200 in the earlier survey.

In the words of one respondent, "Life's tougher than a cane toad that's been flattened by a ute and left to bake on a dusty road."

These figures paint a concerning picture of the nation's emotional well-being.

In a nutshell

  • Surveys are effective when they give a sense of mass opinion.
  • You can cite the quantitative or qualitative results of surveys (i.e. counts & percentages or quotes).
  • Surveys are usually anonymous, so quotes are often unattributed, or attributed to a demographic.
  • Surveys can be badly designed and unrepresentative.