Topic: Media Research Center
MRC Latino's Ken Oliver rants in a Jan. 22 post: "Once again showing systemic anti-Trump bias, the principal national evening newscasts of the six leading Spanish-language television networks in the United States all pointedly declined to report on President Donald Trump’s soaring job approval ratings among Hispanic-Americans."
But that NBC/PBS/Marist poll finding -- putting Hispanic support for Trump at 50 percent -- is questionable at best and very much an outlier; as the Washington Post explains, other polls show Trump's support among Hispanics much lower. Further, only 153 Hispanics were polled, and that particular result has a margin of error of 9.9 percent -- exceptionally wide for a poll, making the result not statistically significant.
Oliver tried to spin the wide margin of error into suggesting that Trump is even more popular among Hispanics, insisting that "the 19-point pro-Trump surge in the Hispanic results" is "substantial even given the 10% margin of error within the subset."
This was followed two days later with a column by Miguel I. Prado gushing over this dubious poll result and declaring that Trump's "approval rating could go on to surpass even this marker by addressing the border crisis at its source. This would be a disruptive, transformational development for all of Latin America."
Olliver then noted in a Feb. 1 post that a couple of the Spanish-language networks noted the poll number -- and also complained that they reported how dubious it is:
The Univision report further attempts to undercut the significance of Trump’s Hispanic boost by pointing out that the 50% level of support registered among Latinos has a 10% margin of error, but the 19% increase far exceeds the margin of error, which means Trump’s actual Hispanic support could be as high as 60% or as low as 40%, which would still be notably higher than the 39% approval rating registered among the U.S. population as a whole.
Way to spin away that ridiculously high margin of error, Ken.