NBC News “Survey Monkey Poll” Adds to Methodological Morass

NBC News, to its credit, has teamed with the Wall Street Journal for a number of years to field national polls conducted by the polling team of Public Opinion Strategies (R) and Hart Research Associates (D). The survey has long been considered one of the best media polls available amid a growing tidal wave of garbage polls with questionable methodologies and even more questionable results.

But NBC News has added to the morass of questionable polling and their methodologies by deciding to release a so-called “Survey Monkey” poll following the first GOP Fox News debate. The “news” was as follows:

“Following Thursday’s Republican debates, an overnight poll shows Donald Trump maintaining his edge over the crowded field of Republican candidates. The survey also finds a majority of Trump supporters say they would vote for him as an independent candidate for president, in the event he does not win the Republican nomination.
The NBC News online poll conducted by SurveyMonkey over Friday night and into Saturday showed possible movement among the candidates following the debates, as well as Republican primary voters declaring Carly Fiorina the overall debate winner.
Donald Trump has the support of 23% of Republican primary voters following the debate, which had the highest viewership ever for a primary debate, with 24 million watchers. The poll was in the field for only a 24-hour period, during which Mr. Trump continued to make more headlines for his negative comments about Fox News moderator Megyn Kelly, which got him disinvited from speaking at a gathering of conservatives in Atlanta.
Ted Cruz and Carly Fiorina were the candidates with the most gains in support, 7 points and 6 points respectively, but there was no real change among most candidates given an error estimate of plus or minus 3.4 % points.”

Survey Monkey, as most know, is an online survey of user-defined respondents, and generally used to gauge member opinion within various organizations and associations, and used for other small bore research projects — not a nationwide survey that purportedly represents public opinion after a key national political event.

In fairness, it’s just one “poll” — and the only way to assess reality is to view surveys in real time as they are released and evaluate cumulative trends.

But this one “poll” and its methodology is weak, clearly stipulating that, Because the sample is based on those who initially self-selected for participation rather than a probability sample, no estimates of sampling error can be calculated.”

Why did NBC News release (and push as “news”) this assemblage of questionable data? Because they could — and because they wanted to be “first” out of the box to make news with branded NBC News polling results.

NBC News marketing types likely, and correctly, decided the inherent, well earned credibility of their regular POS/Hart survey would provide cover for cavalierly flipping out these Survey Monkey results — which ended up widely reported in local news simply because it had the imprimatur of NBC News.

Today’s political polling already has enough problems and issues related to land-lines, cell phones, IVR technique and a variety of other very legitimate factors that have thrown the industry — and its data — into question, contingent upon the actual firm involved.

Unfortunately, NBC News has opted to engage in data slumming in this instance, and the loser, once again, is the public, who hears a quick headline on local TV news and assumes it’s reliable.

 

 

Choose from archives