Election Interference Lurks Where It Is Least Expected
Charles Franklin, Director of the Marquette University Law School Poll, released its latest poll results on April 3. He titled it New Marquette Law School National Survey finds Biden and Trump Each Supported by 50% of Registered Voters. He chose the subtitle Among likely voters Biden draws support of 52%, Trump draws 48%. While some conclusions can be drawn from this survey, none of them have anything to do with whether Trump or Biden will prevail in November or even what the voter preferences are at this moment in time.
This latest poll was conducted between March 18 and 28 using online interviews of people who have agreed to be part of the SSRS (Social Science Research Solutions) Opinion Panel. The SSRS Opinion Panel was established in 2015 and the company says it is “comprised of a nationally representative group of panelists randomly selected throughout the United States”. Members join the panel by invitation and are paid to regularly take surveys on a variety of topics, including politics. The Better Business Bureau has 8 consumer complaints posted online from individuals who say the company would not stop hounding them when they made it clear they were not interested in participating.
The Marquette poll sampled 868 registered voters nationwide. The number of registered voters nationwide is reported at about 161 million, meaning .000539 of 1% of voters were sampled. The margin of error was reported at +/- 4.3 percentage points. 15% of those surveyed said they would vote for someone other than Trump or Biden or stay home, but they were forced to choose one of those two candidates to produce this survey result.
Surveys are a form of research. Credible research is conducted to advance understanding and pursue truth. Intellectual honesty is required to translate the data that are collected into useful information. Credible researchers go to great lengths to utilize valid instruments (The Marquette survey instrument is 22 pages long, and filled with arbitrary questions), eliminate bias, analyze the data, and discuss what can and cannot be concluded from their findings. The limitations of any research must be clearly communicated to prevent over-generalizing, a risk inherent in any research.
The Marquette poll presents the answers to the questions as raw data. There is no substantive discussion concerning the meaning of the results and, most importantly, no guidance concerning its limitations. The overview of the poll results on the Marquette website (https://law.marquette.edu/poll/) is mind-numbing and the multiple tabs available for the detail-oriented could bury an adventurous reader for days.
It doesn’t take an experienced researcher to recognize that this survey is useless as a tool to inform the public about the state of politics and, in particular, the presidential election. The miniscule sample size and sample makeup of people who are willing to take long surveys and interested in being paid for it; the internet format; forcing respondents to make choices they have said they will not make; and the absence of meaningful discussion of survey findings and limitations are among its many flaws.
To illustrate, let’s examine the question that produced the subtitle for the entire poll. Survey respondents were asked if they were registered to vote and about the chances they will vote in the 2024 election. Their choices were absolutely certain, very likely, 50/50, or don’t think you will vote. Certain to vote chose Biden 52% and Trump 48%. Very likely, 50/50, and don’t think you will vote were grouped together and chose Trump by 12 percentage points. Trump 56% and Biden 44%. Not registered chose Trump by 28 percentage points. Trump 64% and Biden 36%.
The researchers leave us with important unanswered questions: Why weren’t the results for each of the choices for chances you will vote reported separately? If the researchers were going to group answers, wouldn’t it have been logical to group very likely to vote with absolutely certain to vote rather than very likely with 50/50 and don’t think you will vote? We have no idea if the not registered group is not registered yet and intends to vote or doesn’t plan to vote. They chose Trump by a huge margin so understanding what not being registered means is significant.
It would be generous to attribute these deficiencies to incompetence but that is hardly the case. The purpose of this survey is to influence. Pollsters are part of the triad exerting a stranglehold on our political process. The political parties and the corrupt media are the other members. Here’s how this works. The survey produces reams of data. In the absence of any disclaimers about what it does or does not mean, the media is free to cherry pick headlines and assign whatever meaning suits their political purposes. The parties, in turn, raise money off of whatever data is useful in advancing their political agendas. Manipulating the perceptions of the public in this way is election interference and it relies on our general laziness when it comes to critically evaluating what we are being told and using our judgment to draw our own conclusions. It is a safe bet that most people will only know of the poll results from what they hear from the media. Few will read the entire survey and fewer still will delve into all of the detail provided. Researchers provide themselves cover by making all of the data available, knowing full well most will not read it.
Reclaiming our country will require each of us to be far more skeptical and rigorous about what we accept as true. Winning is everything in politics and the attempts to influence will be widespread and cloaked in creative ways to deceive. Election interference lurks where it is least expected.