Thread by @jennycohn1

.@GaSecofState study showed that less than 1/2 of voters reviewed machine marked “paper ballots” to confirm the touchscreen voting machines (BMDs) didn’t cheat or err. This is similar to the results of the 2006 study re: DREs & Voter Verifiable Paper Audit Trails. 1/

https://twitter.com/rad_atl/status/1420112306441162754

Here is a link to my article about DREs and Voter Verifiable Paper Audit Trails (VVPATs). It links to the 2006 VVPAT study. 2/ link.medium.com/BRUmGgHofib

Here is a link to the 2006 VVPAT study in which only 40 percent of voters reviewed the VVPAT to confirm the voting machines didn’t cheat or err. 3/ usenix.org/legacy/events/…

Note that neither of these studies assessed the percent of voters who notice and report inaccuracies on the VVPAT or BMD-marked “paper ballot”. Other studies have done so, and the statistics are grim. 4/

In a 2005 MIT study, “Out of 108 elections that contained errors [VVPAT not matching a test voter’s selection], … NO ERRORS WERE REPORTED [by the test voters].” 4/ ted.selker.com/documents/resu…

It is scandalous that vendors went ahead and sold BMDs for use buy most in person voters—and that officials bought them for this purpose—without confirming that voters would do better with them than they had with VVPATs. But that’s exactly what they did. 5/

A study last year by the University of Michigan showed that voters reported only 7% of inaccuracies on their “paper ballot” caused by the voting machine (BMD) in marking it for the voter. 6/ news.engin.umich.edu/2020/01/new-st…

Merely instructing voters to review the printouts didn’t help much. The only thing that did was to have the voters compare the printout to a pre-filled slate, such as a sample ballot. But most voters don’t know to do this. 7/

In other words, although @GASecofState did not study error detection and reporting rates among the 40-50% of voters who review their ballots, earlier studies indicate that those rates would be unacceptably low. 8/

It is unacceptable that @GaSecofState bought these BMDs (despite warnings from experts) without 1st conducting these studies. It is also unacceptable that @GaSecofState did not educate voters to bring completed sample ballots w/ them to the polls to assist w/ verification. 9/

I spent & a few others spent countless unpaid hours trying to educate touchscreen voters (voters who planned to vote on touchscreen voting machines) to use their sample ballots to help verify the accuracy of their “paper ballots.” Why didn’t @GASecofState lend a hand?! 10/

Seriously. @GaSecofState & @GabrielSterling knew about the study from the University of Michigan . Why was I the only one advising voters to use their completed sample ballots to help them verify BMD-marked “paper ballots”???? 11/

Article in post 1 by @markniesse who discovered @GaSecofState’s study via a public records request. #gapol 12/

5/ Correction posts 1-3. The VVPAT study was conducted in 2004, not 2006. It was, however, cited in a 2006 study to which I linked. usenix.org/legacy/events/…

Correction posts 1-3. The VVPAT study was conducted in 2004, not 2006. It was cited in the 2006 study to which I linked. 13/ usenix.org/legacy/events/…