Like micro-breweries, micro-pollsters?

Doug Mataconis posts, Poll: Obama Has 52 Point Lead Over Romney Among Latinos, claiming it as evidence of “the depths to which the popularity of the Republican Party has sunk among Latino voters”.

Bear in mind that the Pew Hispanic Center reports that 23.7 million Latinos are eligible to vote in the USA (myself included).

Now, if you actually look at the poll,

About the Poll
This is the sixth release of an 11-week tracking poll of Latino registered voters. Each week impreMedia and
Latino Decisions will release a new rolling cross-section of 300 completed interviews with Latino registered voters across all 50 states. Battleground interviewers are combined across all six weeks and are 267 completed interviews, with Florida accounting for the largest share of battleground states. Interviews are conducted in
English or Spanish, at the preference of the respondent, all conducted by bilingual interviewers at Latino Decisions calling center, Pacific Market Research. The survey averaged 10 minutes in length, and has an overall margin of error of 5.6% on results that approach a 50/50 distribution. All respondents confirm that
they are Hispanic or Latino and currently registered to vote. This third wave of the survey was fielded Sept 21-Sept 27, 2012

Please correct me if I’m wrong, but I interpret this to mean that they talked to a total of 267 people in 10 states, and after a 10 minute call (the ones at work couldn’t come to the phone), 194 said they prefer Obama over Romney, give or take 10 people.

And that is supposed to be evidence of “the depths to which the popularity of the Republican Party has sunk among Latino voters”?

Meanwhile, a poll with a sample size of 160 people and an 8-point margin of error is brought to you by the Washington Post-ABC pollsters.

Like micro-breweries, micro-pollsters?


For what it’s worth,
Margins of error range from 2% to 4% in national polls, with six out of the seven Real Clear Politics uses on their average staying in the 2% to 3.5% range. An overall margin of error of 5.6% is huge.

Tags: ,

4 Responses to “Like micro-breweries, micro-pollsters?”

  1. Kermit Says:

    For me valid micro poll would mean at the key precinct level and no one is better than that than Karl Rove, no one.

  2. jlh Says:

    Thank you for the exposé. A comment even about a micro-poll, in this media climate, can be dangerous. We live in an MSNBC world.

  3. Fausta Says:

    Mataconis’s commenters are positively apoplectic at my agnosticism.

    Of course, pollsters would never spin, would they?

  4. My friend Pete goes to the diner « The Greenroom Says:

    […] Ingemi – Da Tech Guy – and I were talking about micro-pollsters the other day, and I’m not alone in finding the sample size, em, lacking, to put it midly, I […]