Making sense of polls!
With this year being a presidential election year, we are constantly hearing about public opinion polls! Whether it is which candidate is ahead in national or specific “swing” state contests, Americans' view of the president’s response to the coronavirus, or the public’s prioritization of important problems facing the nation, it can be challenging to know what to make of them all.
Polling is scientific, but also an art. So, take polls with a grain of salt and understand their strengths and limitations. Humans determine which questions should be asked and how, as well as the ways that data should be "crunched.” So, because humans can make errors sometimes, it follows that polls can have shortcomings as well.
In addition, polls cannot always capture subtleties, such as when people might be ashamed to express views or support candidates thought of negatively by larger society. Polls also do not always effectively catch shifts in momentum. They might also fail to predict which people will be motivated to take action on their viewpoints through voting or other means.
Sometimes there is a more fundamental challenge: a mismatch between the question we ask a poll to answer and how a decision is made. For example, in 2016 (and today!), pollsters can ask Americans over and over about their candidate preferences but national percentages will not necessarily represent how electoral college votes will be allocated state-by-state.
So, with all of that in mind, a few tips for reading— and making sense of—public opinion polls:
#1: EVALUATE THE POLLSTER
Is it professional and reliable? No poll is perfect, but we can be more generally confident in findings if the pollster has a past record of well-done polls. Examples might include businesses whose sole work is polling (like Gallup), major news organizations (TV networks, newspapers) or respected scholars at universities. If you have never heard of the pollster, research more before taking results at face value. Even these established pollsters can still get things wrong, so also…
#2: CONSIDER WHO IS BEING POLLED.
How were the polled people found (Were they contacted, and how? Did people just happen upon it online? Were they selected randomly?). It is better for polling to be done in a way where all of the people whose views we are trying to understand have an equal chance of being asked to provide their opinions.
Also, how many people shared their opinions? Was it large enough to give us some faith in their accuracy? Most polls aim for at least 1000 people responding to be reliable. If a poll has a smaller number of people taking it, it tends to be less precise and so we will have less confidence its results clearly stand in for the whole population.
Were lots of types of people (age, race and ethnicity, gender) included? Was it made up of all Americans? Or just registered or likely voters? It is important to be precise when trying to interpret to whom the findings apply. Sometimes we might want to restrict the poll to get a closer look at certain groups or outcomes. But pollsters should ensure they explain who their polling results are meant to represent, and we should aim to understand this as well. And finally…
#3: LOOK FOR TRENDS.
It is possible that a single poll that has results quite distinct from others is "onto something" new or that just has not been caught by other polls yet. But it is usually good to look at several polls or to find a website (Real Clear Politics or Five Thirty Eight, for example) that looks at averages of many polls and takes quality into consideration when it does so.
No matter how good a poll is, it is also important to not over-interpret the polls. They can tell us a great deal about a snapshot in time, but are only able to answer the questions that we ask— and they cannot predict the future We often need to be careful to not take more from them than what we explicitly asked them to do.
And no matter what a poll says, it is probably a good idea to not let it change whether we participate in political decision-making:. The only thing that we may be able to know for certain is that decisions are being made that will affect us, so no matter what the polls say we still must act regardless.
If you liked this post, why not follow me on Instagram so you don’t miss the next one? And, while you are here: check out some of my other recent posts?
And of course, don’t forget, if you have a question about politics and policy I have not yet covered, go to the Contact page and enter it in— I will do my best to write a future post that covers it! Answering your questions is why I am here!
