Posts Tagged ‘statistics’

How Fake News is Made

January 15, 2019

First off, a statement of fact: Donald Trump is a lightning rod. People either really, really love him or really, really despise him. There just aren’t that many folks in the middle. Knowing this, companies, institutions and media organizations on BOTH sides of the political aisle exploit these passionate extremes to great profit.

Donald Trump says and does a lot of [insert adjective(s) here] things that get him a lot of attention, both positive and negative. He’s very good at this and really doesn’t need any help at it. This has not stopped the Hate Industry (or Trump supporters) from fabricating stories about the man so they could cash in on the outrage.

In early 2017, just weeks after his inauguration, a wave of bomb threats were made against Jewish community centers across the US. The Hate Industry and their media stooges swore that this was irrefutable evidence that white supremacists and anti-Semites “were emboldened” by Trump’s election, until it was discovered that the real culprits were a Black man attempting to frame his white ex-girlfriend and an Israeli teenage hacker. Whoosh! The Memory Hole opened and swallowed the story. Nothing to see here.

The Ku Klux Klan was “emboldened” to spray paint “Vote Trump” on the side of a Black church in Jackson, Mississippi, before burning it down, and neo-Nazi arsonists boldly painted “Heil Trump!” on the side of a church in Bloomington, Indiana, before setting it ablaze. Heinous hate crimes that could be traced irrefutably back to President Trump, except for the annoying facts that the first church fire was set by a Black parishioner and the second by its openly gay organist.

Do you recall the avalanche of corrections and retractions that filled the media in the following days? Us neither.

The excellent website, fakehatecrimes.org, has logged over a hundred hate crime hoaxes committed since Election Day, 2016, many of which were explicitly claimed to be Trump-related, until they weren’t.

And the beat goes on. Just last week many “legitimate” media sources breathlessly regurgitated a “report” by two University of Virginia academics who claimed that they had proof that teasing and bullying in middle schools across Virginia had “risen by 9%” in 2017, specifically in areas that had voted for Donald Trump. Did any of the media experts actually examine the report before reprinting the claims? Not so much.

And really, why would they? Even if the story turned out to be inaccurate it’s still solid gold click bait. And when these stories are disproved, they simply go away. No messy cleanup required.

Francis L. Huang and Dewey G. Cornell based their claims on a set of three surveys they have administered to roughly 155,000 seventh- and eighth-grade students across Virginia in 2013, 2015 and 2017, that asks the students their opinions about bullying and teasing in their schools. It’s a worthy effort and the data seems to be as good as any these kinds of instruments produce. It’s worth noting that the results are agnostic, with no sign whatsoever of the terms Obama, Trump, Republican or Democrat to be found.

The problem lies in the way in which Huang and Cornell interpreted the data in their subsequent 2018 report, “School Teasing and Bullying After the Presidential Election.” In that report, the pair claimed an increase in responses to the statement: “Students in this school are teased or put down because of their race or ethnicity,” in the 2017 survey over the previous one. So far, so good.

The duo then take it one step farther by claiming that they have discovered a correlation between the increase in responses and those schools residing in Republican districts. Here is where things start to break down.

Survey results were broken down by eight separate regions across the Commonwealth and numbered 1 through 8. While the percentages of students who answered yes to the “teased because of their race or ethnicity” prompt did increase significantly, the data shows that this is a trend that began during the Obama Administration and shows sizable increases between 2013 and 2015. Where is that “report”?

According to Huang and Cornell’s own data, teasing in the Shenandoah Valley jumped by 63% in President Obama’s second term. Statewide, teasing increased by 35% under POTUS 44. Where’s the outrage over that?

percentages1

In other news, outright bullying, a separate measure from mere teasing, rocketed under President Obama and actually dropped under President Trump, according to Cornell and Huang.

percentages2

Affirmative responses to the prompts “I have been bullied,” “I have been cyber bullied,” and “I have bullied others,” exploded statewide between 2013 and 2015, by 150%, 167% and an unbelievable 220% respectively!

Obviously, it’s pretty easy to cherry-pick data points to put any kind of slant you want into your “report.” Remember the adage about “lies, damned lies, and statistics”? And as we have stated numerous times on this blog, ANY TIME you see a percent symbol in the media, run, do not walk, to see the source documents. And after you’ve done that, follow the money. Who is going to profit from unsubstantiated claims and what consequences do they face if proven wrong?

This isn’t rocket science, folks. Anyone with an Internet connection can do it, providing they want to get at the truth.

Other inconsistencies not addressed in the Huang and Cornell piece include the facts that the three surveys are not identical (some questions are universal to all three, but not all), the 2017 survey includes sixth-graders for the first time, and that the 2017 survey was administered between February and April, when Donald Trump had been in office for less than 100 days, versus Barack Obama’s eight year administration.

If you’re going to blame Trump for teasing then you’re going to have to blame Obama as well.

Lastly, what conclusions did Huang and Cornell actually reach in their study? Despite the hundreds of headlines proclaiming “Trump Causes Teasing!,” the authors themselves concede:

“These findings are correlational and cannot establish a causal relationship but invite the need for further study.”

“The phrase “teasing and bullying” is used broadly in recognition that there are various forms of peer aggression and that student reports do not necessarily conform to an academic definition of bullying.”

“We did not have measures of the quality and intensity of peer aggression or its impact on victims.”

“It is obviously difficult to demonstrate a causal link between statements by a public figure and schoolyard bullying.”

Nobody, friend or foe, can deny that President Donald Trump says a lot of stupid, irrational and sometimes irresponsible things. There is more than enough controversy to go around without cooking up fake news like this malarkey.

Trust, but verify. It’s what used to be known as “journalism” in the old days.

Advertisements

%d bloggers like this: