The following from Mark Hemingway at National Review:
It's All Relative at the L.A. Times
A while back, I did a spit take when I read when the L.A. Times was reporting that voters "narrowly" opposed gay marriage when the survey they were citing indicated voters opposed it by a whopping 19 percent margin.
So yesterday, the L.A. Times reports that an L.A. Times/Bloomberg poll has Obama up 12 percent over McCain. (I haven't looked at the methodology myself, but based on other polls that seems like a significant outlier.)
Anyway, care to take a guess at what word the L.A. Times used to describe that margin? Yep, "sizable."
You see, when it is about the proposition regarding same-sex marriage that will be on the November ballot in California, somehow 54%-35% is narrow. Then if one splits the 11% undecided right down the middle, it becomes 60%-40%. Similar to the margin that Proposition 22 won by in 2000.
But, when it comes to Sen. Messiah Barack a 12% advantage becomes sizable. Oh, and the polling was of registered, not likely, voters. It is known that when polling is done with registered voters versus likely voters, the registered voters sample is tilted to favoring a Democrat because there are more registered Democrats over Republicans. But, likely voters tends to be a more representative sample. Republicans tend to be more in the likely voters catagory. Independents are always the hard group to figure out in polling.
In reality, the problem is that the Left, er LOS Angeles Times is not a serious newspaper. If it were, it would try harder to appeal beyond the left wing reader. It would try to have more serious news, not polls designed as news.
If this continues, readership will continue to hemorrage and maybe then, polling and reporting will be done seriously.