Under the headline “Bluegrass Poll: Grimes takes lead over McConnell,” the Louisville Courier-Journal is out with a STORY this afternoon that reads as follows:

After two polls in his favor, Senate Minority Leader Mitch McConnell has slipped behind Democratic challenger Alison Lundergan Grimes, according to the latest Bluegrass Poll.

Grimes, Kentucky’s secretary of state, now leads the veteran five-term senator 46 percent to 44 percent among likely voters, the survey found. While that advantage is within the poll’s margin of error, it represents a 6-point swing to the Democrat since the survey was last conducted in late August.

I’m sorry, but that headline and story are wrong on several counts. Even though the text acknowledges that Grimes’ “advantage is within the poll’s margin of error,” the conclusion that she has taken a lead is not supported by the survey. The writer apparently doesn’t know how to read polls.

I’m not saying that Grimes isn’t actually ahead of McConnell. For all I know, she might well be. I’m just saying that the poll cited by the Courier-Journal doesn’t support that conclusion.

You see, the problem is that the writer misinterpreted the significance of the poll’s stated “margin of error” of plus or minus 4 percent.

Here’s a good rule of thumb in interpreting poll results in a race between two candidates: If Candidate A’s lead over Candidate B is not at least twice as large as the stated margin of error, it should be considered insignificant.

In the example from the Louisville paper, Grimes’ lead of 2 percentage points was not twice as large as the margin of error of 4 percentage points. So, the two candidates could be said to be in a close race. And since the difference between their respective numbers is only two points, it’s fair to say that they seem to be in a statistical tie.

Considering the margin of error of 4 percentage points, the 46-percent figure accorded to Grimes could, in reality, be 4 points higher or 4 points lower. That also goes for the 44-percent figure accorded to McConnell.

I don’t want to get too technical about this, but a margin of error in a well-conducted poll usually relates to a 95-percent “confidence factor.” What that means is that if you conducted the same poll 100 times in the same time frame, using the same methods, the results in 95 of those polls would be within the stated margin of error.

Granted, this is all green-eyeshade stuff, but that’s the way professional polling works.

The bottom line: If you see a story about a poll purporting to show that Smith is ahead of Jones by, say, 3 or 4 percentage points, it’s best to consider them virtually tied. Remember, a lead that’s less than twice the margin of error isn’t much of a lead at all.

One other thing: Polls that are conducted by unscientific means include no margin of error and are utterly worthless.

Under the headline “Bluegrass Poll: Grimes takes lead over McConnell,” the Louisville Courier-Journal is out with a STORY this afternoon that reads as follows:

After two polls in his favor, Senate Minority Leader Mitch McConnell has slipped behind Democratic challenger Alison Lundergan Grimes, according to the latest Bluegrass Poll.

Grimes, Kentucky’s secretary of state, now leads the veteran five-term senator 46 percent to 44 percent among likely voters, the survey found. While that advantage is within the poll’s margin of error, it represents a 6-point swing to the Democrat since the survey was last conducted in late August.

I’m sorry, but that headline and story are wrong on several counts. Even though the text acknowledges that Grimes’ “advantage is within the poll’s margin of error,” the conclusion that she has taken a lead is not supported by the survey. The writer apparently doesn’t know how to read polls.

I’m not saying that Grimes isn’t actually ahead of McConnell. For all I know, she might well be. I’m just saying that the poll cited by the Courier-Journal doesn’t support that conclusion.

You see, the problem is that the writer misinterpreted the significance of the poll’s stated “margin of error” of plus or minus 4 percent.

Here’s a good rule of thumb in interpreting poll results in a race between two candidates: If Candidate A’s lead over Candidate B is not at least twice as large as the stated margin of error, it should be considered insignificant.

In the example from the Louisville paper, Grimes’ lead of 2 percentage points was not twice as large as the margin of error of 4 percentage points. So, the two candidates could be said to be in a close race. And since the difference between their respective numbers is only two points, it’s fair to say that they seem to be in a statistical tie.

Considering the margin of error of 4 percentage points, the 46-percent figure accorded to Grimes could, in reality, be 4 points higher or 4 points lower. That also goes for the 44-percent figure accorded to McConnell.

I don’t want to get too technical about this, but a margin of error in a well-conducted poll usually relates to a 95-percent “confidence factor.” What that means is that if you conducted the same poll 100 times in the same time frame, using the same methods, the results in 95 of those polls would be within the stated margin of error.

Granted, this is all green-eyeshade stuff, but that’s the way professional polling works.

The bottom line: If you see a story about a poll purporting to show that Smith is ahead of Jones by, say, 3 or 4 percentage points, it’s best to consider them virtually tied. Remember, a lead that’s less than twice the margin of error isn’t much of a lead at all.

One other thing: Polls that are conducted by unscientific means include no margin of error and are utterly worthless.