The Campaign Spot

Election-driven news and views . . . by Jim Geraghty.

The Washington Post Has a Sample-Size Problem


See, Washington Post, this is why a lot of people don’t trust your polls. The big headline in today’s story is, “52 percent of likely voters across swing states side with Obama and 41 percent with Romney in the new national poll.”

Only later today, after their own Jen Rubin calls them out, do they clarify, “Out of 929 registered voters in the new poll, 161 live in one of these eight states, with a margin of sampling error of eight points.”

The margin of error for most national polls we see is two to four points; in most state polls, the margin of error tops at about 4.5 percent or so. This one has a margin of error of eight points. What’s more, if some pollster tried to use a sample of 161 respondents for a state poll, most political watchers would scoff and dismiss it as too small a sample to give a useful measurement.

The Washington Post takes a sample of 161 respondents in eight states, and uses that figure to run a dramatic story.

Tags: Polling


Subscribe to National Review