It's a Very Small Sample of 200 People, But the One Day Track of the NBC Poll Taken After the Debate Shows Hillary's Lead Down to 2

Please do note all the caveats packed in there -- this was supposed to be a poll taken over several days. The number of respondents for this one day is very small, and would come with a big margin of error.

In addition, if this is just a fragment of the full poll, then they wouldn't have adjusted/weighed the numbers.* Which is a critical step, even though people seem to think it's devious. You mostly get older people and people on landlines in polls, and you do have to correct for the fact you're oversampling that particular population, and undersampling younger people and those without landlines.

To be honest, after all these caveats and cautions, I don't know why I'm even posting this.

Oh that's right -- because all other polling news is pretty miserable but human beings have a built-in survival instinct to seek a pathway of hope.

Anyway, in the one day track, Hillary is only up over Trump by two points post debate.

According to the survey data collected by NBC/WSJ on October 10, the only post-debate day for which data are available, Clinton’s lead in two-way race is only two percentage points. In a four-way race, the post-debate data shows a Clinton lead of five percentage points.

The October 10 survey data can be ascertained by comparing the NBC/WSJ poll data from the pre-debate surveys on October 8 and 9 to the aggregated poll results from surveys conducted on October 8, 9, and 10.

That is, by comparing NBC's early-release numbers with the final numbers, Sean Davis has backed-out the numbers for the last day of polling.

Hey, it's somethin'.

* Actually, given that these numbers were weighed before inclusion in the final poll, I don't know if this caveat applies.

It should be noted, however, that NBC released its own partial numbers early, just to drive the "Trump collapses" narrative they engineered.

posted by Ace at 03:54 PM

| Access Comments