Cover image from Real Clear Politics
Within the news media, one surefire way to flesh out political stories is by using poll data. Given human nature, it’s a pretty good way to indicate to the people who always want to be on a winning side, which side is actually doing the winning. (In the world of sports this is known as being a fair weather fan.) In the current presidential election year, polling data has become an issue unto itself that is used to either raise up or tear down a single candidate.
Thus it was a shock to the system when the polling data started coming in from the state of Wisconsin, a largely blue collar place where Democrats tend to have control, that the Republican front runner was not ahead in the polls when his largest support is coming from betrayed Democrats and the silent majority of independents. In fact, the polls were significantly different than the national trend, enough so as to turn heads and prompt rejoicing in other candidates’ camps. Having been educated in statistics and learning how to put together poll questions and sample sizes to get results that could possibly reflect a pre-determined reality, this writer – and many others like her – had serious doubts that the polls were really telling the whole story.
As it happens, one of the most ballyhooed polls to be released last week, the Marquette University Law School Poll, may very well have compromised. In a very much ignored report, the Washington Examiner tells us of a hacking, of sorts, that happened before the results of the poll were released.
While the poll was a state-wide poll, and did have a wide demographic base of respondents, 31% of the respondents to the poll live within the city limits of Milwaukee, Wisconsin and the surrounding suburbs. A lot of confusion accompanied the release of the polling data, and the presidential portion of the poll was leaked (somehow) to media outlets prior to the official release of the data. Sources within Marquette’s law school claimed that computer systems were accessed remotely without the approval of the school by “various parties”, and the school said it is reviewing its policies to determine how to prevent such a leak in the future.
That, of course, renders the results of the Marquette poll questionable at best. As it happens, that poll had a sample size of less than 500 Republicans and did indeed show the nationally second place candidate as having a double digit lead in the state, which contradicts subsequent polls, especially among independents. Other polls, despite the reporting, are more reflective of what we have seen so far this election season. Does this look familiar?
On April 5 in Wisconsin, the reality will be fleshed out, and unless the voting machines do their magic tricks all over the state of turning in 100% votes in any one precinct to a single candidate, the state of the presidential election will be told definitively. At least for now.
For more analysis and information on polling and how the information is presented in a way that may not reflect the thoughts of the people visit this link.