CT: UConn Memory Card Reports: More Info, More “Junk” Data

Results also indicate continuing extensive problems with procedures not being followed by election officials.

The UConn VoTeR Center recently released two more reports in its series of memory card audits <Aug Primary Post-Election> <November Pre-Election>

We applaud UConn for carrying out these tests and the Secretary of the State’s Office for commissioning these tests. We appreciate the valuable addition of analysis of the Audit Logs.

From the Presidential Pre-Election Report:

The UConn VoTeR Center performed a pre-election audit of the memory cards for the Accu-Vote Optical Scan tabulators that are to be used in the November 2008 Elections in Connecticut. The cards are programmed by LHS Associates of Methuen, Massachusetts, and shipped to Connecticut districts for use in the elections. For the pre-election audit the VoTeR Center received and examined 620 memory cards as of November 3, 2008. These cards correspond to 620 distinct districts in Connecticut. About 2/3 of these memory cards were randomly chosen by the VoTeR Center personnel during the visits to LHS and before the cards were packed and shipped to the towns. Another 1/3 of the memory cards came from the towns directly, where the cards were randomly chosen for pre- election audit (this procedure applied to the town for which the cards were not selected at LHS). The towns that shipped the cards to be audited were not asked to perform the local pre-election testing before shipping, thus the majority of the cards received from the towns did not contain the pre-election testing events in their audit logs.

This document reports on the findings obtained during the audit. Among the 620 cards received and tested as of this writing, 564 cards (91%) were found to have been properly programmed for election. These cards contained valid ballot data and the executable code on these cards was the expected code, with no extraneous data or code.

One of the ongoing areas of concern is the quality of memory cards and the quality control of our vendor LHS Associates, which supplies the Diebold memory cards and programs them for each election. Past tests at UConn, in other states, and surveys by the audit Coalition indicate significant problems.

The source of the problems remain in question: Are the cards bad from the beginning? Do they leave LHS working correctly, to be harmed by transport or by testing by election officials? These latest reports may get us closer to some of the answers. Looking at the results and the reports conclusions, we conclude:

  1. The quality control is getting worse: The pre-election test showed 3.8% of the cards were programmed correctly but contained additional random data, apparently a new issue.
  2. Election officials continue to violate procedures: 16 of 297 post-election cards sent by registrars were from the wrong election. 23 cards were duplicated which is prohibited by the Secretary of the State’s procedures.
  3. The cards are bad when they leave LHS: 8.9% of the cards tested post-election contained “Junk” data, when 2/3 of the cards tested were directly from LHS. UConn noted no significant difference in card failure rates between those collected at LHS and those shipped by Election Officials

We might conclude, perhaps erroneously that:

  • The problems are getting worse: The August Post-Election report showed 15.4% of cards with “Junk Data”, while previous post-election reports have shown 5% and 8% junk data.

However, the memory cards shipped by election officials are not selected at random and do not represent all the districts in the election. The results may reflect a tendency for election officials to send in the “Junk” data cards or for districts with problems not to send cards – we have no way of knowing.

The results show serious problems with the hardware and the election system: Memory cards should not fail at these rates. In the unlikely event the memory card problems are shipping and handling, then LHS should specify exactly how the shipping and handling should be improved. Election officials should be expected to follow procedures. How can the public have confidence in our elections when procedures are not followed?

We note from the pre-election report:

The current SOTS procedure is not to duplicate cards, but this is now being reconsidered, with one possibility being to allow duplication at the districts and to document and report all such occurrences.

We have reservations about making this change and question the added vulnerability associated with the duplication of cards. We have concerns that election officials will follow any such procedure given that they have clearly failed to follow the current procedure for memory cards and a similar procedure associated with the post-election audit requiring them to call the Secretary of the State’s Office when incorrectly sealed ballots are discovered. Procedures are insufficient when they are difficult to follow, they are regularly violated, they cannot be effectively monitored, are not enforced, and are not enforceable.


Facebooktwittergoogle_plusredditpinterestlinkedintumblrmailFacebooktwittergoogle_plusredditpinterestlinkedintumblrmail

Leave a Reply

You must be logged in to post a comment.