When poll numbers tell us nothing

Screen Shot 2017-08-02 at 16.53.50 PM
Click to enlarge if you want.

In the wake of Trump’s momentous tweet last month, those with an anti-transgender agenda pointed to polls demonstrating that Americans agree with the president’s claim that the military would be better off without transgender people serving in it.

Those poll NUMBERS indeed seem to support that statement. One poll conducted in late June showed that 31 percent of Americans said it was bad for the military to have transgender people serving. Only 21 percent said it was good and 38 percent said they had no idea.

Another poll, much more widely, shared said 44 percent of “likely voters” agreed with Trump’s tweet, while 45 percent disagree, showing the country pretty evenly split.

Let’s go back to college and talk about polling and survey methods. If you recall, your professor harped on question formation. How you formulate your question means everything in survey research.  The questions you ask, if they aren’t considered thoughtfully, will result in your survey measuring the wrong opinion.

Let’s go back and look at those polls cited above. The first one asked:

“The U.S. Department of Defense now allows transgender people, those who identify with and want to live as the opposite sex, to serve openly in the military. Is this decision good for the military, bad for the military or does it have no impact?”

The majority of Americans have no educated opinion on this matter! How on earth does the average American know anything about this? They don’t! it’s the wrong question because people’s opinion on this matter is irrelevant: they aren’t qualified to have an opinion.

The other poll asked survey respondents to “agree”, “disagree” or select “undecided” to the following statement, which you should recognize because it’s what Trump tweeted:

“Our military must be focused on decisive and overwhelming victory and cannot be burdened with the tremendous medical costs and disruption that transgender in the military would entail.”

Now this is a more “valid” question because anyone can have an opinion about what someone says or does. A person needs no qualifications to judge this. But this is still not the right question! Who cares if people agree with the president? This is not a presidential issue, it’s a policy issue. You have to ask the right question.

On July 26 Reuters released polling results which tell a completely different story. Reuter’s poll shows that nearly two-thirds of Americans believe transgender people should be in the military. How could this poll show such an overwhelming majority of people supporting transgender service people when the other two did not?

It asked a different question:

“Transgender people should be allowed to serve in the military.”

Only 27 percent of people responded “they should not be allowed” and the rest of respondents didn’t know how to answer the question. That latter group is likely more qualified to have that opinion because most people in America don’t know what it’s like to serve in the military because they haven’t served themselves.

It’s interesting how numbers can always be used to argue for a specific agenda. That’s why it’s so important to look beyond the numbers. For clues always exist which clarify just what the numbers are saying. In this case, what they’re saying is: you gotta ask the right question.