How Texas pollsters do their jobs when everyone screens their calls

It’s election season. There will likely be a new political poll every week – if not every day – until November. And as a careful news consumer, you may even notice polls released just days apart that give significantly different results about which candidate is leading in a particular race and by how much.

Why is that?

“Pollsters are not predictors, they don’t look into a crystal ball,” said Brandon Rottinghaus, a political science professor who conducts polls at the University of Houston.

Rather, Rottinghaus said that pollsters are “just giving us a sense of what the snapshot of the electorate looks like at this moment, based on the factors they have identified as relevant to their poll.”

It’s important to keep the “factors they found relevant” part in mind, especially when looking at surveys that specifically target specific populations.

“It’s never best to look at national polls to figure out how a subgroup is doing because the sample just isn’t big enough,” NPR senior political editor Domenico Montanaro told The Texas Newsroom.

Take, for example, recent polls that show former President Donald Trump could win by a 10% margin quarter of black votesor assumptions made from surveys Young black men are becoming more conservative.

Pollsters interviewed for this article said black voters are the hardest to gauge, largely because they don’t respond to polls in large numbers.

“For example, we have difficulty reaching a significant percentage of Latino respondents. In some communities, it is also difficult to attract African-American respondents,” Rottinghouse said. “These are definitely problems when it comes to drawing conclusions about populations where they may have very specific opinions.”

One group of Americans that may be easier to survey are “older, white people,” said Joshua Blank, research director at UT Austin’s Texas Politics Project, which conducts a variety of polls on politics and trending topics.

Blank said sometimes when they do surveys they get something too many Respondents in this population group. The pollsters then have to change the math.

“Let’s say there were 200 old people in the survey and it should have been 100. Well, then instead we count each of their answers as 0.5 – half – of an answer. And that’s how we get them up to weight,” Blank said.

“Similarly, with young African-American men, it might say, ‘Hey, we should have had 30 in the survey, but we only have 15. Well, instead of counting them as one each, let’s count them as two.’ ‘,” he added.

Who do the pollsters survey?

Ideally, people conducting surveys could simply increase the sample size of their respondents to get more accurate results.

In the example above, you may have wondered why pollsters don’t just go out and talk to 15 more people. It turns out that this is a lot harder these days than it used to be.

Richard Murray is a senior research fellow at the University of Houston’s Hobby School of Public Affairs and has been conducting surveys since the 1970s, what he calls the “golden age of surveying.”

Back then, people had landlines and didn’t monitor their calls like they do today.

“You could buy a sample of randomly generated numbers, whether they were listed or not. That was a pretty accurate sample of the entire population,” he said.

As technology advanced, so did the cost of conducting an accurate political poll.

“Typically today, a telephone survey will probably cost you $100 per completed interview. If it’s done over the internet, maybe half that amount,” Murray said. “As with all things in life, cost is important.”

You read that correctly. The effort to interview a single Texan for a survey today could cost $100.

“For convenience, if you want to do a really meaningful survey from a well-known agency, you’re probably going to spend between $40 and $60,000 for a sample of about 1,500 people,” Rottinghaus said.

(Actually in a current one Think about a job interviewTime magazine’s Philip Elliot estimates that it would cost “nearly a million dollars to conduct a statewide survey in Texas that would give us an accurate count for every county.”

This price means that large, well-funded organizations are typically the ones with the best surveys.

The frontrunners include the New York Times and Sienna College poll. They are Rated #1 from FiveThirtyEight, a website focused on opinion poll analysis. They rate pollsters based on how close their results are to the outcome of an actual election.

Murray said another reason The New York Times ranks so highly is that it has continued to conduct polls traditionally.

“The New York Times still calls people at home, but they had to make hundreds of calls to get a single interview. And for most of us, that doesn’t work. “That’s not possible, I don’t have the money,” said Murray.

Overall, the pollsters interviewed for this article recommend approaching surveys with caution.

The next time you see a sensational story about a new survey on the television screen (or hear about it on the radio), they encourage you to pay attention to the number of respondents, the margin of error, and the organization that conducted the survey.

“What it comes down to is, is it easy for you to figure out what they did and how they did it? That’s how it should be,” Blank said. “And if not, I think you have to immediately ask questions about the quality of the work or the intent behind it.”