LA Times Polling Averages Distorted Toward Trump by One 19 Year Old Man in Illinois

[louis c]

<snip> A lone 19-year-old black man in Illinois is single-handedly (and unknowingly) changing the standings of the presidential election, The New York Times has found. Due to the 19-year-old Trump supporter's inordinate weight in a USC Dornsife/Los Angeles Times Daybreak panel, his poll answers are directly affecting national averages, like the one compiled by RealClearPolitics. <snip>


Link:
http://theweek.com/speedreads/654767/how-single-19yearold-man-illinois-wildly-skewed-major-poll-towards-trump

Link to NY Times:
http://www.nytimes.com/2016/10/13/upshot/how-one-19-year-old-illinois-man-is-distorting-national-polling-averages.html?_r=1

[Cicada]

1. Similarly, Rassmussen showed Trump getting 24% of black vote

I guess a sampling error can lead to big errors with small groups.

[louis c]

3. The thing is..........

..........that this LA Times poll does not "refresh" the respondents. They are the same 3,000 people, over and over again. This single individual is weighted is so many categories which have few participants, that he is counted and weighted over and over again, during every poll he responds to. This methodology is substantially flawed and should not be counted in any average.

[bluestateguy]

2. Then why doesn't Silver and RCP just junk that poll?

Stop putting it in their averages. Just behave as if it does not exist.

[oberliner]

4. That is one weird survey

Props to them for being transparent - but it is sort of nonsensical.