Robert Mueller in 2012 (Credit: White House/Wikipedia)

A new CBS News poll on the Robert Mueller investigation illustrates why many people feel polls can’t be trusted to reflect public opinion.

According to that poll, only 44% of Americans feel the “investigation into dealings between Trump associates and Russia” is justified. Yet, 76% think Trump should cooperate with that investigation, and 64% think Republicans in Congress should let the investigation continue.

How much sense does that make? If most people don’t think the investigation is justified, then why would three-quarters of them think Trump should cooperate, and two-thirds think the investigation should continue?

One problem with this series of questions is that the original question about the investigation is confusing.

“As you may know, there is an investigation into dealings between Trump associates and Russia. Do you think the investigation is justified, or is the investigation politically motivated?”

For some inexplicable reason, the pollsters decided not to ask the very simple question – “Do you think the investigation is justified or not?”

Instead, they complicated the response by suggesting the opposite of “justified” is “politically motivated.” The problem is that some people may see that the investigation is both justified and politically motivated.

In fact, part of the confusion may be that “politically motivated” may mean different things to different people.

Although the pollsters implied that an investigation that was politically motivated would not be justified, the two subsequent questions in the poll suggests that many respondents had no problem with that description.

The second question in the series asked:

Should Donald Trump cooperate if he is asked to be interviewed as part of the Russia investigation?

According to the results of this question, even a majority of Republicans believe Trump should agree to be interviewed, along with three-quarters of independents, and nine out of ten Democrats.

Given that the poll shows, for example, that only 11% of Republicans say the investigation is justified, one can’t help but be mystified that the same poll shows 53% wanting Trump to cooperate with the investigation.

The last question in the series makes any literal interpretation of the results problematic.

“Which of these should the Republicans in Congress do about the Russia investigation? Should they 1) try to end the investigation, or 2) let the investigation continue?”

If we take the results of this series of questions literally, that means that 11% of Republicans think the investigation is justified, but only 29% want it to continue – while a majority think Trump should cooperate with it.

Democrats are somewhat more consistent: 72% think the investigation is justified, 91% want Trump to cooperate with it, and 90% feel it should continue. Still, even here, many more Democrats think Trump should cooperate with the investigation, and that it should continue, than think it is justified.

And independents also reveal largely conflicting views: Only 42% says the investigation is justified, but 76% think Trump should cooperate with it, and 65% think it should continue.

Top-of-Mind Polling on Mueller Investigation

Apart from the confusing wording of the initial question, the poll results suffer from not measuring how strongly people feel about the issue. It’s highly unlikely that 95% of the public has thought carefully about the issues involved in the investigation, yet that is what the poll results suggest.

For many respondents, the answers they give are top-of-mind, and not necessarily consistent with each other – the responses influenced by the wording of the question, and the order in which questions are asked.

Many respondents who have not thought about the issue are nevertheless pressured by the “forced-choice” format of the question (no explicit “unsure” option provided) to come up with an answer. But that answer won’t necessarily be consistent with other answers they give.

Had the pollsters asked respondents whether they felt strongly or not strongly about their views, we would have a better understanding of how many Americans have actually given some serious thought to the issues, and how many have only a vague awareness. And the poll results among respondents with “strong” views would almost certainly have been more consistent among the three questions posed than what was reported.

Instead, without any measure of how many Americans have actually given much thought to the issues, the poll gives us conflicting and inconsistent results that confuse more than they enlighten.

CLARIFICATION

The original version of this story said 44% of Republicans think the investigation is justified; it is 11%. We regret the error.