[E-voting] Study: Two-Thirds of Voters Fail to Notice Vote-Flipping on Touch-Screen DRE Voting Systems

[Catherine Ansbro]

http://www.bradblog.com/?p=4682
[Interesting recently released doctoral thesis shows that voters don't 
verify or notice errors when using electronic voting machines. See the 
original article for live links, including link to the doctoral thesis 
itself.]

*Study: Two-Thirds of Voters Fail to Notice Vote-Flipping on 
Touch-Screen DRE Voting Systems
*
If Voters Don't Notice Incorrect Votes on The Review Screen in Front of 
Their Face, Why Would They Notice it on a So-Called 'Paper Trail' 
Printed Afterward?

/Guest Blogged by John Gideon of VotersUnite.org 
<http://www.VotersUnite.org>/

A new study supports the contentions of Election Integrity advocates who 
oppose Rush Holt's election reform bill <http://www.BradBlog.com/Holt> 
on the basis that it allows for the use of Direct Recording Electronic 
(DRE) touch-screen voting systems. Supporters of the bill claim that the 
bills mandate to require so-called "Voter-Verified Paper Audit Trail" 
(VVPAT) printers on such DRE systems mitigate the concerns about the 
machines notorious inaccuracy and proven tamperability. The new study 
strongly indicates otherwise.

In her recently released doctoral thesis, “The Usability of Electronic 
Voting Machines and How Votes Can Be Changed Without Detection” [PDF] 
<http://chil.rice.edu/research/pdf/EverettDissertation.pdf>, Rice 
University researcher and Doctoral candidate Sarah P. Everett reveals 
that review screens, presented to voters at the end of the voting 
process on DRE voting machines, fail to be effective.

Of 66 participants from the general Houston population, with a median 
age of 45 and an even distribution of educational levels *only 32% of 
the voters noticed that malicious changes had been made to their review 
screens during mock election testing*.

A further test was accomplished with 101 participants, with an even 
gender split (51 males, 50 females), a median age of 40, with most 
having either some college education or holding a degree. Of those, 
*only 37% of the participants noticed that _vote flipping_ had occurred 
on their review screens during mock election testing*.

Also shocking is that of those 101 participants 6% walked away from the 
voting machine without pushing the button to cast their ballot. Instead, 
they just left the voting process at the final review screen. Thus, had 
this been a real election, a full 6% of the voters would /not/ have had 
their ballots counted at all.

So when only about one-third of the voters actually review their ballot 
on the ballot review screen and notice problems why would anyone think 
that adding a "Voter-Verified Paper Audit Trail" printer to a DRE will 
solve any problem? Rather than adding useless and expensive printers; 
why not just get rid of these machines all together?

In fact, Ms. Everett states, when it comes the use of DRE voting systems 
[/emphasis added/]...

[A]s the situation currently stands, voters cannot be depended upon to 
check the validity of their vote. Many security experts and election 
reform groups are calling for VVPATs to be required on all DREs and as 
of the 2006 elections, nearly half of the states mandated that their 
DREs have paper trails (electionline.org, 2006). However, these studies 
show that solutions to DRE security problems that require voter 
verification of their ballots may not solve vote-flipping problems. 
*Users are not even checking their ballots on the review screen that is 
presented directly in front of them.*

===