Misrepresenting the Failure of the Democratic Party

In a Washington Post article written by Aaron Blake on April 3rd, 2018 they allege that fake news might have won Donald Trump the 2016 election. An interesting claim in an election where the orange man was the one claiming fake news was being used to work against him throughout the entire election. Of course, a "distinguished" journalistic organization like the Washington Post would only make such a claim on the back of significant data right? I decided to dig in to see!

Source: Pixabay

What does their study say?

Their study asked 281 questions via online survey that contained 3 fake news statements that had been spread during the campaign cycle that were negative about Hillary. They found that for those who had previously voted for Obama in 2012, 89% who did not believe any of the fake news to be true voted for Hillary in 2016. When they believed 1 of the fake news items, 61% voted for Hillary. Finally, for those who believed 2 or 3 fake news stories, only 17% voted for Hillary. Here were the 3 fake news stories they peppered in their survey.

• "Hillary Clinton is in very poor health due to a serious illness"
• "Pope Francis endorsed Donald Trump for president prior to the election"
• "During her time as U.S. Secretary of State, Hillary Clinton approved weapon sales to Islamic jihadists, including ISIS"

Their conclusion when trying to factor in all the data was that approximately 4 percent of Obama's previous supporters did not vote for Hillary because of fake news stories, which could have been significant enough to swing the election the other way. So, essentially, fake news led to Hillary's defeat.

Issues outlined in the article/research used

The Study Was Not Peer-Reviewed

Source: Pixabay

Here's what the Washington Post had to say about the study:

"The study from researchers at Ohio State University finds that fake news probably played a significant role in depressing Hillary Clinton's support on Election Day. The study, which has not been peer-reviewed but which may be the first look at how fake news affected voter choices, suggests that about 4 percent of President Barack Obama's 2012 supporters were dissuaded from voting for Clinton in 2016 by belief in fake news stories."
Source: Washington Post

Typically, you peer-review studies like this so expert analysts can verify that you have statistically significant data and that the methods used to analyze the data are accurate and truthful. In this case, instead of going with a peer-reviewed approach, they just made their conclusions and posted their findings publicly. To the credit of the Washington Post, they at least mentioned this upfront instead of making me discover this. So, in order to analyze this further we have to operate under the assumption that the metrics they're presenting are accurate since we don't have the underlying data or authentication via a peer-review process. So, let's move on assuming it's accurate I guess...

Based on 585 respondents
The conclusion is based on the 585 respondents among a 1,600 total sample size. For scale, that's 0.88% analyzed responses when compared to the 66,444 students enrolled in Ohio State University where the study was done (2.4% for all answers collected). This seems like a somewhat small sample size when compared with Ohio State University, much less the sample of all voters in Ohio, or all voters in swing states, or all 2012 Barack Obama voters, or all voters in general. Yet, they claim it should be representative of the entire nation. Their rationalization for this is that they paired down the data from 1,600 to 585 respondents based on a nationally representative sample of gender, age, race, education, ideology, and political interest.

"This well-established propensity-score-matching technique yielded a nationally representative sample"
Source: Ohio State University Study

Okay, so that's great if we have a statistically significant set of data to start with, but do we have the sample size needed for this conclusion? If their claim is that about 4% of Obama's 2012 supporters were turned away because of fake news stories, we can calculate the confidence interval by using the population of Obama supporters in 2012 and the sample size (leaving the confidence level at 95% and the percentage at the standard 50%). I went ahead and used a calculator to determine this based on the population size being 65,915,795 and the sample size being the 585 used records for the analysis. The confidence interval is 4.05 based on this calculation. This means that they show between -0.05% and 8.05% of Obama voters could have been pushed away from Clinton by fake news.

If they wanted to be able to determine with 95% accuracy within 1 percent the impact of the news, they really needed to have a sample size of 9,603 responses. If they wanted to get 99% accuracy within 1 percent they'd need 16,637. They just didn't quite hit the mark on a statistically significant sample size to prove their point here. Perhaps this is the reason they didn't seek to get this study peer-reviewed?

Source: Pixabay

Other factors they entertained
Don't worry, these guys did understand there might be other factors besides fake news in the survey. Here's some of the other possibilities they entertained.

"The Clinton campaign heavily emphasized gender-related issues in an attempt to mobilize female voters. Could this have alienated men to the extent that they abandoned their 2012 support for the Democratic presidential candidate?"
Source: Ohio State University Study

Wait, did they rip that right out of Hillary's "What Happened" book or something? Seems like one of her poorly thought out excuses. Regardless, they determined that this was not the case by analyzing male vs female numbers represented in their data. What they couldn't see in their data is it's possible that emphasizing gender-related (perceived) issues might have turned off both independent females and males who aren't firmly in the democratic echo-chamber. But, let's assume there's nothing to see here.

"Did the absence of an African-American presidential candidate from the top of the Democratic ticket lead black voters to waiver in their commitment to the Democratic candidate?"
Source: Ohio State University Study

Ah, if it's not a gender thing it must be a race thing right! Again, they look into their data and find that black voters were slightly less likely to switch votes than white voters indicating this was not the reason. They also go on to find that age and education are not significant contributing factors either.

Finally, they find a few things that could correlate. People who were ideologically more conservative defected half of the time compared to 14% who were on the left. 39% of those who thought the economy was in poor or very poor shape defected compared to 12% who thought it was good or very good. Party affiliation among the previous Obama supporters also was significant, showing that 7% of democrats defected compared to 40% of independents and 68% who identified as Republican, Libertarian, or Green parties. These actually sound pretty reasonable as contributing factors. Let's flip back to the Washington Post article aaaaaand they're gone. No mention of the other factors the study thought might contribute in public facing article. I'm sure all your readers dive into the data to find that information though so that's cool, no worries about the omission of data here!

Are the questions fair?

First of all, let's look at their 3 questions. 1 is a very generic questioning of Hillary's health implying she has a "serious illness"... I think there's been plenty of reason to be suspicious that her health might not be great. She seemed to have an awfully hard time walking down some stairs in India that nobody else seemed to have an issue with recently.

I would venture to say several people that read that question might not have heard the "fake news" they're referring to, but might recall how she's appeared on camera a few times and thought to themselves, "oh that makes a lot of sense, that's probably true." Seems questionable to attribute this to be the reason some didn't vote for her.

For the second question regarding Pope Francis. I sincerely hope that people don't base their voting decisions based on what someone who is not living in America says anyway, but perhaps if you look at the approximately 24% of Americans that are Catholics and then take out the Democrats that voted for Obama in the previous cycle that did see that specific news story before they voted and then pick out those of that group who base their entire voting philosophy around what their pope says I guess it could have had some impact. Doesn't seem like were generating election swinging numbers here though.

For the third question about selling weapons to ISIS. Well... when WikiLeaks confirmed that some of her weapons dealings eventually fell in the hands of ISIS I don't blame people for falling for the semantics being slightly different in the question. This was even validated by Snopes, who I often feel try to put a liberal spin on their fact checking. This seems like a leading question to get the results they wanted and it was their most successful in tricking people. If you can attribute some of Obama's voter base for ditching Hillary because they thought she was involved in direct sales of weapons to ISIS, do you think you're turning them back to Hillary by clarifying that she didn't sell the weapons "directly" to them but they still did end up with the weapons?

Summary

So, outside of the fact that they didn't have a statistically sufficient sample size and that they also recognized some other factors with a similar level of impact on results outside of their 3 questions but then proceeded to not show any of those in the article, and that their 3 questions are quite questionable to start with, this all seems totally legit right? Oh wait, that's right, there's tons of other reasons people might vote differently in one election over another, like maybe that she's a blatant liar riddled with scandals (Benghazi, Libya, Emails, etc.)...

Let's call this what it is, leftist denial and failure to accept the faults of the Democratic party. They ran their literal worst option against arguably the Republicans worst option and lost against a man who can't even control his Twitter urges. They lost the evangelical christian vote overwhelmingly to a known philanderer. For those saying this is reflective of a corruption in Christianity, you're reading into this incorrectly. The Christian community didn't all the sudden become vastly less virtuous over a 4 year period. This means that the Christian community actually thought Trump was of higher moral character (or that his policies were at least) than your candidate... ouch. Democrats ran a scandalous, deceitful, and callous woman that called those who disagree with her deplorables, that didn't feel the need to bestow many battleground areas with her almighty presence, that used identity politics to mask that she didn't know what the hell she was talking about, and that had an excuse for everything. Perhaps the Washington Post could run another article for their Democratic readers that is more brutally honest and helpful to their party. Try this headline:

Hillary was a shit candidate that actually deserved to lose to an orange twitter troll

I didn't even support Trump in this election and I really hate that we ended up with the two main party options we did, but the election is over. Stop crying. Stop with the excuses. Just... stop already. Swallow the pill that is reality and I challenge you to put a legitimate candidate in the race next time.

Generated with imgflip template

If you enjoy this content, please consider upvoting/following/resteeming.

This post is unofficially brought to you by Chaco Canyon Gold ale from Empyrean Brewing Co. here in my home state of Nebraska. If you want to encourage more content like this, consider a donation. Creating content is easiest when paired with the smooth finish of a well crafted beer, and beer is most delicious when it is generously funded by those who enjoy good content!

Bitcoin (BTC): 15r2ktpJ8Tkw3MDc5GUmYEZBbBWSdv4r9p
Ethereum (ETH): 0x290a784EA092DDB9e665Bc8268eE242b3Fc269D1
Litecoin (LTC): LPJLDRqP9TdCPyQxPWuFbSFUi3ZJmAgrHK