Since the Sept. 11 terrorist attacks, most large media outlets have conducted polls to gauge public opinion on how the U.S. should respond to the tragedy.

Almost all the news coverage of these polls emphasizes that huge majorities of Americans favor war. This is clearly true. But it’s also true that, judging from the polling conducted by the major media outlets, there is significant opposition among U.S. citizens to the use of military force in the Persian Gulf.

For example, consider a recent New York Times story stating that 68 percent of Americans support using military force, even if it means killing thousands of civilians. Support is slightly higher among men and slightly lower among women.

Imagine if these polling results had been reported this way in the news media: Just after the worst terrorist attack on the U.S. in history, a third of the American public opposes engaging in war if thousands of innocent civilians are killed.

Or what if the news reports interpreted the poll like this: After over 5,000 U.S. civilians were killed in terrorist attacks on September 11, 40 percent of women reject the idea that America should respond with military force, if thousands of innocent civilians are killed …. It’s amazing, isn’t it, that right after this horrible act, such a huge number of people oppose war? Isn’t this the real news?

Apparently not, because not a single major news outlet framed its poll results this way. Instead, the dominant message that emerged from the news media’s reporting of the polls was that Americans are hungering, even impatient, for war.

Another popular question in polls was whether U.S. citizens would support increases for military spending and intelligence-gathering efforts to fight terrorism. News outlets reported that over two-thirds of Americans favored increases for intelligence gathering or for the military. And one-third supported increased spending, even if the government had to cut social programs or use the Social Security surplus.

But, again, what if the results from the Los Angeles Times poll had been reported like this: In the weeks after the destruction of the World Trade Center by terrorists, half of Americans opposed increasing the intelligence budgets if it means cutting social programs and dipping into the Social Security surplus.

Usually critiques of polls focus on the questions presented. Pollsters will say that the results of polls depend on which questions are asked, and how they are worded. But recent polls illustrate another problem: how you describe the results. It’s the responsibility of journalists to describe polling data from various perspectives, to give citizens a full understanding of its possible meaning.

Other startling polling data that was woefully under-interpreted after Sept. 11 was this: over 70 percent of Americans acknowledged that economic hardship in Muslim countries, created by western capitalism, was part of what motivated the terrorist attack on the United States.

This public sentiment, illustrating a complex understanding of the motivations for terrorism, is inconsistent with a simplistic emphasis on a military response to the tragedy. Overall, the American people’s views on going to war to fight terrorism are much more complex than the news reporting of its own polling data.

Unfortunately, it seems that reporters at major news outlets had made assumptions about what they expected the public to think, prior to conducting their polls. Once the polling data was in, they simply wrote stories confirming their assumptions – which are now widely believed.

Yet, it turns out that a lot of citizens – even if it isn’t a majority – don’t want our nation to kill innocent civilians and are more interested in understanding what’s behind the terrorist attacks. That’s important news that should be investigated and reported in more detail.

Jason Salzman is board chair of Rocky Mountain Media Watch, a nonprofit organization challenging the news media to meet the basic standards of professional journalism. E-mail him at