Those new Robert Mueller numbers don’t mean what you think they mean

Polls. Sometimes we love them, and sometimes we don’t. How do we know when they are accurate? Sometimes we don’t. A valid poll can display feelings on issues. However, if the poll is biased or the questions unclear, results don’t always speak the truth.

This week, USA Today published an article entitled “Poll: Half of Americans say Trump is victim of a ‘witch hunt’ as trust in Mueller erodes.” As you might expect, “President” Donald Trump immediately tweeted in support of this headline, without reading the full article. If you take this headline at face value without reading the article and conducting further research, you might think that confidence in Mueller has dropped significantly since the December NPR/PBS NewHouse/Marist poll that found only 33% of Americans viewed Mueller’s investigation as a “witch hunt.”

As reported by Vox, the Suffolk/USA Today poll question which generated the headline is “seriously flawed.” What does that mean? The question was a two-part question, which can result in skewed numbers from lack of clarity. The question asked: “President Trump has called the Special Counsel’s investigation a ‘witch hunt’ and said he’s been subjected to more investigations than previous presidents because of politics. Do you agree?”

Some may agree with the first part of the question but not the second or vice versa. The two parts are not necessarily related to or dependent upon each other, and different answers could be given for each. In today’s fast-paced world, most of us “skim” materials to save time and could have easily missed one part of this question. As Vox points out, this is poor polling strategy.

Though neither Vox nor I have proof of any of this, the compound question invariably sets itself up for failure, especially viewed in light of similar but well-constructed poll questions. Vox points to a Washington Post / Schar School poll conducted last month, which indicated that 56% of those polled find Mueller more credible than Trump, and the Hill.TV / American Barometer poll found 58% who believe Mueller’s investigation is “unbiased.” Vox has this right. When presented with one clear question, people participating in polls gave very different responses.

Vox does mention other results of the Suffolk/USA poll that are based on more clearly written questions and tend to support the belief that the earlier-mentioned results were skewed, giving Trump some bad news in the process: 62% of respondents are against introducing impeachment proceedings at the present time while also supporting the aggressive investigation techniques being used by Democrats. Nancy Pelosi is right to play down impeachment—for now. Her decision is in line with the American public’s view. Of course, Trump chose to ignore the worst news for him from this poll: 52% have “’little or no trust’” in his denial of collusion. We can’t always take poll results at face value and certainly not by a headline, which could also be erroneously constructed.

Wisely take poll data with the proverbial grain of salt. Many contingencies come into play that can easily skew a poll in the direction intended by the polltaker or framer. Of course, Donald Trump sees anything that appears to support him as good, without bothering to look further. The headline plays to his base, and that’s what he wants.