Put this poll in your pipe and smoke it

Conservative bloggers have been peeing in their pants lately with delight over polls that show great public dissatisfaction with the Democratic-controlled Congress.

They ignore, of course, the fact that the low numbers for Congress reflect, in part, great dismay that lawmakers aren't taking a harder stance against the Bush administration. The right-wingers want us to think that the polls reflect an anti-liberal bent among the public, which is nonsense.

And the conservatives have got nothing to say about a Pew poll showing that Americans favor the Democratic Party over the Republicans by a margin of 15 percentage points. Karl Rove's dreams of a permanent Republican majority are gone with the wind, as this piece notes.

The media can't read polls

The Rascal doesn't give a rat's rectum about Sunday's presidential election in France, but it annoys me to no end that the global media are misreading the latest poll results on the race.

As you can see here, Reuters is out with a story stating that one candidate is "well ahead" of the other. The story calls it a "commanding lead." But the poll of 1,415 people actually shows a margin of just 54 percent to 46 percent. That is not a commanding lead. It is not necessarily a lead of 8 percentage points.

If the margin of error in the poll is, say, plus or minus 3 percentage points, which seems reasonable with a sample of that size, the race might well be extremely close.

The news media regularly misread poll results. Reporters and editors generally are ignorant of the most basic principles of public-opinion polling. The public knows even less about such matters.

We'll have more to say about this issue as next year's U.S. presidential election draws nearer.