Web Developer Says Donald Trump’s Media Survey Might Really Be Fake News

Those who have been amusing themselves by gaming Donald Trump’s media survey might be wasting their time. According to at least one web developer, it’s quite possibly “fake news” in a very real way. He believes the site may not actually be recording input at all, merely presenting a narrative that will encourage Trump supporters to donate.

As previously reported, the survey, which was described as a questionnaire to determine how Americans feel about media honesty, included questions that many read as very agenda-driven and leading.

Many who opposed Trump began experimenting with the survey, learning that it could also be accessed by people outside the U.S., and that it could be done multiple times.

In fact, Trump actually released two separate media surveys, explaining to subscribers in an email that the first one had been “attacked” by Democrats who took the survey to try to sabotage the results.

Apparently too many people "sabotaged" Trump's media survey by…taking it. So he scrapped the results/started over. https://t.co/XVfR2ZWbwN pic.twitter.com/34iduFMCSp

— Mel Gillman (@melgillman) February 22, 2017

It’s not clear what the results were in either poll, but after completing the survey, the user is taken to a page asking for donations.

NPR assessed the poll early on, declaring it “phenomenally biased,” but “also useful.” Here, the poll is described as committing “a variety of polling sins” including using leading questions, relying on incomplete data, and being directed at a self-selected group for biased results.

However, NPR posited that the media survey was designed to help Donald Trump in a different way: leading questions that would plant ideas in a reader’s mind.

“Those biased questions can plant ideas in people’s heads. If Jane Doe happens to be a Trump supporter and a Christian, she might not think that she’s being particularly hurt by news organizations — until she sees a question like, ‘Do you believe that people of faith have been unfairly characterized by the media?'”

Of course, this has no bearing on those who see Donald Trump as a man with a fragile ego, who they hope to thwart by hurting his feelings with unfavorable survey responses. These voters dove in, ready to prevent Trump — whose administration, according to Deadline, just this weekend, prevented press access to several networks that Trump has derided as “fake news” — from using the results of the media survey as a basis for further action.

For these survey-takers, the results were the point. They were determined to create results that would neither boost the president’s self-esteem, nor drive an anti-media agenda.

https://twitter.com/artangeIs/status/832461977561608192

Just took this Trump media accountability survey to let them know that I value a critical and vocal free press. https://t.co/S9pGODHqNP

— Stuart Fraser (@speculatifraser) February 17, 2017

However, Devin Pike on Hacker Noon spoke to a web developer who said that, as far as he could tell, the media survey isn’t actually recording responses. Instead, the developer described it as a “fraud” that existed only to raise money, not to collect input.

Pike gets a lot more technical, but the essence is that the web developer looked into the site’s code while filling out the survey, and found that it didn’t appear to actually record the submitted data.

Following up with other coders, Pike got mixed reviews, with several reporting the same findings, and others sending screenshots that appeared to show the data transmitting to the server normally. Yet another reported that he saw data transmitting to the server, but not being recorded. While the theory was posited that perhaps the coding had been messed up by an honest accident when moving the survey from the GOP website to the Donald J. Trump campaign site, yet another developer told Pike it had worked fine for him on both websites.

The final conclusion regarding Donald Trump’s media survey? It may or may not be recording the data users are filling in. At least four coders or developers say that it doesn’t appear to do so — and others disagree.

[Featured Image by Olivier Douliery-Pool/Getty Images]