Claim: A 2016 Stanford investigate suggested widespread primary choosing rascal in mixed states bearing Hillary Clinton.
WHAT’S TRUE: Two researchers (presumably connoisseur students) from Stanford University and Tilburg University co-authored a paper reporting they unclosed information suggesting widespread primary choosing rascal bearing Hillary Clinton had occurred opposite mixed states.
WHAT’S FALSE: The paper was not a “Stanford Study,” and a authors concurred their claims and investigate methodology had not been theme to any form of counterpart examination or educational scrutiny.
Example: [Collected around e-mail, Facebook, and Twitter, Jun 2016]
Origin:On 8 Jun 2016, a Facebook page “The Bern Report” common a request authored by researchers Axel Geijsel of Tilburg University in The Netherlands and Rodolfo Cortes Barragan of Stanford University suggesting that “the outcomes of a 2016 Democratic Party assignment competition [are not] totally legitimate:
That amicable media share described a document as “a illusory investigate square put together by a integrate of college students, Rodolfo Cortes Barragan Axel Geijsel.” That request (properly termed a “paper,” not a “study,” as a latter tenure implies some form of veteran vetting) resolved with a matter that a information examined by a author “suggest that choosing rascal is occurring in a 2016 Democratic Party Presidential Primary election” and that “this rascal has overwhelmingly benefited Secretary Clinton during a responsibility of Senator Sanders”:
Are a formula we are witnessing in a 2016 primary elections trustworthy? While Donald Trump enjoyed a transparent and early corner over his Republican rivals, a Democratic competition between former Secretary of State Hillary Clinton and Senator Bernard Sanders has been distant some-more competitive. At present, Secretary Clinton enjoys an apparent advantage over Sanders. Is this claimed advantage legitimate? We contend that it is not, and advise an reason for a advantage: States that are during risk for choosing rascal in 2016 evenly and overwhelmingly preference Secretary Clinton. We yield concentration justification for this claim.
First, we uncover that it is probable to detect irregularities in a 2016 Democratic Primaries by comparing a states that have tough paper justification of all a placed votes to states that do not have this tough paper evidence. Second, we examination a final formula in 2016 to a unbecoming exit polls. Furthermore, we uncover that no such irregularities occurred in a 2008 rival choosing cycle involving Secretary Clinton opposite President Obama. As such, we find that in states wherein voting rascal has a top intensity to occur, systematic efforts competence have taken place to yield Secretary Clinton with an farfetched domain of support.
In an appendix, Geijsel and Barragan settled that their investigate was still in swell and had not nonetheless been theme to counterpart review, nonetheless given a information was rarely accepted they believed it improved to pre-release their commentary due to a ongoing primary list count in California (among other factors):
Statement on peer-review: We note that this essay has not been strictly peer-reviewed in a systematic biography yet. Doing so will take us several months. As such, given a timeliness of a topic, we motionless to tell on a Bern Report after we perceived rough certain feedback from dual professors (both experts in a quantitative amicable sciences). We devise on seeking peer-reviewed announcement during a after time. As of now, we know there competence be errors in some numbers (one has been identified and sent to us: it was a mislabeling). We enthuse anyone to let us know if they find any other error. Our aim here truly is to know a patterns of results, and to enthuse others to rivet with a electoral system.
The post-introduction apportionment of a paper began with a comparison of outcomes in “primary states with paper trails and nonetheless paper trails,” holding that potentially false formula led a researchers to “restrict [our] research to a proxy: a commission of representatives won by Secretary Clinton and Senator Sanders.” After identifying around a Ballotpedia web site 18 states that use a form of paper corroboration for votes compared to 13 states nonetheless such a “paper trail,” they resolved that states nonetheless “paper trails” demonstrated a aloft rate of support for Hillary Clinton:
Analysis: The [data] uncover a statistically poignant disproportion between a groups. States nonetheless paper trails yielded aloft support for Secretary Clinton than states with paper trails. As such, a intensity for choosing rascal in voting procedures is strongly compared to extended electoral outcomes for Secretary Clinton. In a Appendix, we uncover that this attribute binds even above and over choice explanations, including a prevalent domestic beliefs and a changes in support over time.
The information enclosed in a Appendix didn’t series accurately what those choice explanations competence be:
Are there other variables that could criticism for a categorical outcome (states nonetheless paper trails going overwhelmingly for Clinton)? We conducted a retrogression indication and enclosed a % of Non-Hispanic Whites in a state as of a final Census, a state’s electoral story from 1992 to 2012 of bearing Democratic or Republican nominees for President (i.e., a “blueness” of a state), and a non-static of interest: paper route vs. no paper trail. As expected, race/ethnicity and domestic beliefs played a role: The Whiter and some-more magnanimous a state, a reduction it adored Clinton. However, a outcome for paper route stays significant. States with paper trails uncover significantly reduction support for Clinton. As such, even over a intensity for other approaching factors to play a role, a intensity for rascal is compared with gains for Clinton.
Dependent variable: Percent support for Clinton in a primaries
In a paper’s second portion, a researchers examined discrepancies between exit polls and final formula by state, a theme of discuss (hashtagged #ExitPollGate on amicable media) that antedated a announcement of their paper and was addressed in a Nation essay encountering a explain that exit polls suggested fraud. The Nation‘s research hold that rascal showing exit polling sundry significantly from a form of exit polling typically carried out in a United States:
While exit polls are used to detect intensity rascal in some countries, ours aren’t designed, and aren’t accurate enough, to accomplish that purpose. [A polling association VP], who has conducted exit polls in frail democracies like Ukraine and Venezuela, explained that there are 3 essential differences between their exit polls and a own. Polls designed to detect rascal rest on interviews with many some-more people during many some-more polling places, and they use really brief questionnaires, mostly with usually one or dual questions, since ours customarily have twenty or more. Shorter questionnaires lead to aloft response rates. Higher response rates interconnected with incomparable samples outcome in many smaller margins of error. They’re distant some-more precise. But it costs a lot some-more to control that kind of survey, and a media companies that unite a exit polls are usually meddlesome in providing provender for pundits and TV articulate heads. All they wish to know is that groups came out to opinion and why, so that’s what they compensate for.
Contrarily, Geijsel and Barragan settled in their paper that:
Anomalies exist between exit polls and final results
Data procurement: We performed exit check information from a database kept by an consultant on a American elections.
Analysis: On a overall, are a exit polls opposite from a final results? Yes they are. The information uncover reduce support for Secretary Clinton in exit polls than a final formula would suggest. While an outcome distance of 0.71 is utterly substantial, and suggests a substantial disproportion between exit polls and outcomes, we approaching that this disproportion would be even some-more farfetched in states nonetheless paper voting trails. Indeed, a outcome distance in states nonetheless paper voting trails is extremely larger: 1.50, and yields some-more farfetched support for a Secretary in a hours following a exit polls.
The consultant whose numbers were employed for a paper wasn’t specifically cited by name, nonetheless his moniker seemed on a associated spreadsheet: Richard Charnin. Charnin indeed lists some considerable statistical certification on his personal blog, nonetheless he also appears to spend many of his concentration on conspiracy theories compared to a JFK assassination (which raises a doubt of either his math skills outstrip his ability to request doubtful logic to data).
Geijsel addressed questions about exit check numbers in a successive e-mail to a blogger who was rarely skeptical of his research:
In short, exit polling works regulating a domain of error, we will always design it to be rather off a final result. This is mostly mentioned as being a domain of error, mostly put during 95%, it indicates that there’s a 95% possibility that a final outcome will distortion within this margin. In exit polling this is mostly distributed as fibbing around 3%. The bigger a difference, a smaller a possibility that a outcome is legitimate. This is since nonetheless those exit polls are not 100% accurate, they’re accurate adequate to use them as a anxiety point. In contrariety to a thought that substantially 1 out of 20 formula will differ. Our formula showed that (relatively) a outrageous volume of states differed. This would lead to dual possibilities, a) a Sanders supporters are FAR some-more peaceful to take a exit polls, or b) there is choosing rascal during play.
Considering a context of these sold elections, we trust it’s a latter. Though that’s a personal opinion, and others competence differ in that, we trust we can successfully disagree for that in a private environment deliberation a weight of a possess study, a beliefs of other statisticians who have both looked during a possess investigate (and who have conducted corroborating studies), and a fact that a internet is dirty with tough justification of both voter termination and choosing rascal carrying taken place.
That blogger upheld a anlysis on to his father (“a late Professor Emeritus in Mathematics and Applied Statistics during a University of Northern Colorado”), Donald T. Searls, Ph.D., for comment:
I simply asked him to examination it in full and send me his comments as to a methodology and his perspective as to a validity. For a record, he has been a Republican for as prolonged as we can remember and has no seductiveness in voting for a Democratic nominee, whoever that competence be. we perceived his response around e-mail today. Here is what he wrote:
I like a research really many adult to a indicate of requesting luck theory. we consider a information pronounce for itself (themselves). It is always cryptic to request luck speculation to experimental data. Theoretically different confounding factors could be present. The tender information is in my mind really absolute and transparent on a own.
My personal opinion is that a whole routine has been fraudulent opposite Bernie during each turn and that is harmful even nonetheless we don’t determine with him.
I called him after receiving his response to [ask him to] explain his remarks on a focus of luck speculation to a data. His criticism to me was that he did not trust it was required for a authors to take that step. If he had finished a investigate himself, he would not have worried with doing so. As he said, a information speaks for itself.
Although Geijsel cited a series of sources to justify a explain that rascal was well-documented in a 2016 primary season, many of those citations concerned persons with an seductiveness in a altogether brawl (such as groups celebration to lawsuits). That cause doesn’t indispensably expel doubt on a researchers’ findings, nonetheless it highlights that not many eccentric and neutral corroboration of their conclusions has occurred yet.
Last updated: 15 Jun 2016
Originally published: 15 Jun 2016
Geijsel, Axel and Rodolfo Cortes Barragan. “Are We Witnessing a Dishonest Election?”
7 Jun 2016.
Holland, Joshua. “Reminder: Exit-Poll Conspiracy Theories Are Totally Baseless.”
The Nation. 7 Jun 2016.
Booman Tribune. “My Dad’s View of Election Fraud Study.”
11 Jun 2016.
Booman Tribune. “Election Fraud Study Authors Respond.”
13 Jun 2016.