CNN launches new voting standards as 2020 race for competition gets under way

The news is flooded with polls every year of elections. Those looking to raise their profile have found that frequent polls on hot topics deserve attention. But all of these surveys do not deserve to be taken as seriously as those conducted by researchers conducting surveys to help the public understand the American people.

The standards are based on responses to 16 questions assessing the methodology and assumptions upon which any publicly-reported survey is based. The pollsters who meet the standards are the ones who answered the 16 questions and who did it in a way that shows that they are using industry best practices and proven methodologies.

It will be easy to answer CNN's 16 questions if a sounder has taken the necessary steps to reliably assess the opinion of the American public. Any pollster who can not or will not answer all may not have put the time and effort necessary to reach a representative sample today.

And getting the views of a representative group of Americans is more difficult than ever. The polls changed a lot and quickly. For decades, the gold standard of polls was to investigate people affected by a random sample of telephone numbers for landline phones. Today, this approach alone produces biased results, as the majority of Americans who only use cell phones are ignored. But calling mobile phones as well as landlines has become more expensive and more difficult. Online research has become dominant in market research and is gaining prominence in public opinion polls as telephone surveys have led to increased costs, even though there was no clear consensus on best way to proceed.

Probabilistic sampling – the type where each member of the group of respondents has a chance to be interviewed – remains the best place to start in survey methodology. This is always the rule of gold and research based on other methods is lost without established benchmarks with the help of these methods. But this is not the only way to get a representative share of the American public.

By implementing these new standards, CNN will cover some surveys using non-probability sampling, but only if investigators can demonstrate that their modeling assumptions have been carefully developed using criteria relevant to the subject and that their samples were chosen from reliable sources. target universe.

This is where the list of CNN's 16 questions comes in. Evaluating the validity of a survey is only possible if pollsters are transparent about what they do. Transparency has become an essential part of how respondents rate the validity of other searches. The American Association for Public Opinion Research and the Roper Center have both implemented formal means to evaluate the transparency of a sounder. CNN's new voting standards mean that readers and viewers can be sure that CNN's polls are done by investigators ready to show their work.

There is not a good way to answer the 16 questions that make up our standards questionnaire, but some things do not meet CNN's reporting standards and will not be. And there are some clear new lines that define how polls should be done today. We will not cover:

We take seriously our responsibility to report surveys carefully.

When reporters ask themselves whether or not they should report information shared by a source, one of the most important considerations is how this source knows what they say they know. A journalist must be convinced that the information from the source is reliable before distributing it to the public.

Our assessment of polls and surveys should follow the same standard, and the initial approach should be skeptical: the existence of a survey is not enough to make it new and journalists should not be more credulous about numbers than about words. CNN's standards mean that surveys covered by CNN have been thoroughly verified.