Cooking the heds with Fox
The poll finds that nearly half of Americans agree with Petraeus that some progress is being made under the surge: 49 percent think the increase in troops has led to improvements in Iraq (17 percent "major" improvements and 32 percent "minor"), while 45 percent say the surge has not made much of a difference.
Giving rise to two points:
1) Never infer answers to questions you haven't asked. "Has the surge led to improvements in Iraq" doesn't and can't produce an answer to "Did the surge work?"
2) Don't, um, lie about the percentages. The sum of those agreeing on major or minor improvements is 49 percent, plausibly summarized in the inside hed as "nearly half." Calculating the confidence interval for both those responses at a 95% confidence level, a result accurately reflecting the population could lie anywhere between about 43.5% (adding the lower end of the range for "minor" and "major") and 54.5% (adding the upper ends). If you can get "most" to think "surge worked" out of that, a career in high-stakes gambling is not for you.
The next case, though, is just a nice basic lesson in how to fail your first hed-writing assignment (and I hope we're listening out there in radioland):
Remember all those droning lessons from your undergraduate days? That a killing isn't a "murder" until there's a murder conviction. This is such. As the story notes, the perp was convicted of voluntary manslaughter. Her husband was killed gruesomely, violently and deliberately, but he wasn't "murdered." "Killing" or "slaying" would have worked fine; "death" would be correct but misleading. But we can't say his "murder" was a tragic event because there was no murder.
Inept or biased? Fox reports, you decide!