International polling provides information about how people in different countries are thinking about issues like immigration, technology, religion, you name it. But polling in different parts of the world can be very challenging, because what works in one country may not work in a different country.

For this episode of Methods 101, I’m enlisting the help of some of my colleagues who grapple with those challenges every day.

The first step in conducting an international poll is determining where it’s possible to conduct rigorous survey research. In some countries, governments insist on approving the questionnaire, thus potentially limiting what can be asked.

In other cases, laws may discourage public opinion research altogether. Also, armed conflicts or political unrest can make it too dangerous to survey in certain parts of the world, but keep in mind, it’s not just the political or security environment that matters.

Polling is a relatively new enterprise in some parts of the world, so there may not be many local partners who have the experience and knowhow to conduct high-quality surveys.

The bottom line is that researchers have to carefully consider whether the necessary research infrastructure and skills are available on the ground, whether conditions are safe for field work, and whether government regulations make it possible to ask the kinds of questions
researchers need to ask.

So after pollsters decide which countries to survey they need to determine how to administer the interview. In some countries, nearly everyone has internet access, making online polling feasible. In other parts of the world, telephone usage is higher so telephone surveys are more effective for reaching a representative cross section of the public.

But, in countries where many people are not easily reached by the internet or phone, in person or face-to-face interviewing is often used. However, this type of polling comes with its own set of challenges. For example, pollsters conducting face-to-face surveys have to think carefully about where to send each interviewer in order to reach a diverse set of the population while balancing the time and money it requires to travel across a country.

Also, in some countries, nearly all households have addresses making it more straight-forward to draw a national sample. But in other countries, many households don’t have an address, which poses a challenge for interviewers in the fields.

Fortunately, thanks to technology, field interviewers are increasingly provided with GPS-enabled tablets, smartphones, and even satellite imagery to help them conduct random sampling in locations where addresses or up-to-date maps are not available.

It’s also important to determine which languages are spoken in a country so that researchers can translate the questionnaire when needed. Here at the Pew Research Center, for example, we translated our 2018 Global Attitudes Survey into more than 40 languages, including 11 languages in India alone.

Translation raises a whole other set of challenges around consistency though, as researchers want questions to be comparable in terms of meaning, no matter what language they’re administered in. For instance, to translate the phrase “a free society,” you want to ensure that “free” is translated to mean “open” or “unrestrained” as opposed to without cost.

It’s also important to avoid idioms or phrases that may be common in one country but not in another. And whenever possible, it’s a good strategy to have local experts review questions for any cultural or political sensitivities to be aware of. If resources allow, testing the survey in all the languages will help identify whether a translation is confusing or likely to be misinterpreted.

So, as you can see, there is no one-size-fits-all approach to conducting an international poll. Success depends on addressing the technological, societal, and cultural dynamics unique to each country, and working with professionals who are trained and local to the country is a critical part in making that happen.