UConn Memory Card Report: More garbage in, some good information out

Last week, the University of Connecticut UConn released its latest memory card report:  Technological Audit of Memory Cards for the April 24, 2012 Connecticut Primary Elections <report>

We can easily echo our summary of the previous report.

We applaud Dr. Alexander Shvartsman and his team for developing the technology to perform these innovative tests, the diligence to perform the tedious tests, and the fortitude to report the facts.

We do not applaud the lack of cooperation of officials in the audit or the lack of official compliance with memory card procedures. We are left wondering if this is the level of compliance and cooperation when officials know their efforts will be disclosed: “What is their compliance when their actions are unlikely or impossible to scrutinize?” Can you imagine such numbers from any other technology or Government function? Where is the outrage?

Take a look at these statistics for the election with 598 districts each expected to send in a card before and after the elections:

Prior to the primary 110 out of 598 districts sent cards, that is 18.5% compliance

After the primary 105 out of 598 districts sent cards, that is 17.6% compliance,

however, only 49 of those cards were used in the election, a compliance rate of 8.2%

UConn expressed doubts, as we also have, that the cards were actually selected randomly as directed.

UConn concluded that there is a need for compliance with directives and procedures, not only in the rate of sending in cards, but in following election procedures:

We make the following concluding remarks and recommendations.

The SOTS Office should continue publicizing proper procedures and continue offering training. In particular, to reinforce the need to prepare all cards for election prior to the election dayand prior to the pre-election audit.

Fewer cards are being duplicated at the districts, and it is important to continue reiterating that cards must never be duplicated. Any cases of duplication should recorded in the moderators’ logs and be brought to the attention of the SOTS Office with a documented explanation of why this is necessary.

It is important for the districts to report any problems during pre-election testing (and any card problems) to the SOTS Oce as soon as possible upon completion of the tests.

It is important for the districts report to the SOTS Office any unexpected behavior of the tabulators that seem to necessitate a restart or a memory card reset. It would be helpful if moderators’ logs contained records of machine restarts, perceived causes, and reasoning for the restart or reset. There was at least one documented case of a tabulator malfunction during  this primary election. In such cases it is strongly recommended that the problematic tabulator is tested by the Center personnel (either at the district or in our laboratory).

The current number of cards with unreadable data (junk data) continues to be high. We have determined that weak batteries are the primary cause of this. The vendor developed a new non-volatile, battery-less memory card, and our ongoing evaluation continues to con rm their compatibility with the AV-OS machines used Connecticut. A limited pilot using the new cards was successfully performed in Vernon. It is expected that a broader pilot deployment of the new cards by the SOTS Oce will occur in the near future. The use of the new card should eliminate the major cause of memory card failures.

It is important that cards sent for the pre-election audit are selected at random. One card randomly selected from four cards in each district is to be randomly selected for the audit. While the districts are encouraged to submit all malfunctioning cards to VoTeR Center, all such cards need to be identi ed separately from the cards randomly selected for the audit. When a suciently large collection of cards is selected randomly for audit, the results of the audit meaningfully represent the overall State landscape and help identify technological and procedural problems that need to be solved. Should the selection not be at random, for example, by avoiding sending duplicated cards in for audit, the results are less representative, and may lead to masking technological problems. Therefore training should continue stressing the need  to submit appropriate cards for the pre-election audit.

For the post-election we received fewer than expected number of cards, 155, out of which only 49 were used in the election. This is a very low number. It would be extremely important in the future to obtain substantially larger numbers of cards from the actual use in the elections.

It is indeed good news that their has been a successful first test of new memory cards. Hopefully, further testing will be successful and will result in a relatively speedy full deployment:

New non-volatile (battery-less) memory card was recently developed by the vendor. Our preliminary analysis of this card con cermed that it is compatible with AV-OS systems deployed in Connecticut. A pilot deployment of the new cards was done in the Town of Vernon using 12 of the new cards. The cards performed well, no failures were detected, and no such cards lost their data. However this is a very small sample of cards. We are currently performing in-depth testing of the non-volatile cards and as of this writing the results are encouraging.

A broader pilot is being planned by the SOTS Oce to occur in the near future. The use of the new card should eliminate the major cause of memory card failures.

Facebooktwittergoogle_plusredditpinterestlinkedintumblrmailFacebooktwittergoogle_plusredditpinterestlinkedintumblrmail

Leave a Reply

You must be logged in to post a comment.