A straw poll helps you understand what the public thinks about a given issue or person. It’s an unofficial vote that can be used to help people to make more effective decisions. A poll on voting intention, for example, might help politicians understand why some voters are opting for their rival—and change what they say during the election campaign. There are two types of polls: “scientific”, that use random sampling controls to get a sample that’s statistically representative of a population, and “unscientific” polls that are cast out without any sampling controls. In both cases, they provide quick results for making snappy decisions. In large groups, spontaneous straw polls are used to gauge the amount of support for a particular idea.

It depends how you use them. Your biggest nemesis is selection bias. In the 1936 US presidential election, The Literary Digest sent a poll to people in motor vehicle registries and telephone books, and believed the result was an accurate representation of the voting population. It wasn’t. During the Great Depression, many Americans couldn’t afford cars and phones, and as a result they weren’t consulted in the poll. Think about how to poll an authentic cross-section of the population. Evaluate the profiles of people that use your main distribution channels, and figure out how to reach people outside those channels. Most importantly, you need the right tool. Typeform’s interface is human-friendly on desktop and mobile. Plus it’s effective—people engage more, leading to more responses.

Think carefully about your question. Aim to ask something in the most unbiased way possible. Words sway people—for example, in the 2014 Scottish independence referendum the original question: “Do you agree that Scotland should be an independent country?” was changed to: “Should Scotland be an independent country?” Why? Because polling experts concluded that “do you agree” frames the question in a positive light—people don’t really like disagreeing. Even for straw polls, you may also experience “first-listing bias” where the first candidate in the list gains a handful more votes than the second candidate. With Typeform's poll maker you can randomize the answers to eliminate the chance of that happening.

Think about the demographic of your voters—the presentation of your poll needs to cater to all of them. Choose the clearest font possible, and a color scheme that’s friendly to people with color blindness. A lot of voters won’t be aware who the candidates are—use Typeform’s Image question type to solve this problem. Decide whether you want to display photos of the candidates or their party’s logo. When you make a straw poll, think short and sweet. The more data the better, and you wouldn’t want to put off potential voters by asking for their life story. One or two follow-up questions could really help your campaign—you might, for example, want to ask about the issues that matter most to your respondents. Ask, listen, act.