[Date Prev][Date Next] [Chronological] [Thread] [Top]

RE: AVTS - Provisional Voters - Numbers



From: owner-support@gesn.com [mailto:owner-support@gesn.com]On Behalf Of Steve Knecht
Sent: Thursday, February 01, 2001 4:10 PM
To: Global Support
Subject: AVTS - Provisional Voters - Numbers

There was a question regarding how often the pollworker would be dealing with provisional voters in a precinct based election.  I asked Elaine, Brad, and Charles at Alameda yesterday.  The consensus was that now they would see the highest number being about 60 in some Berekeley precincts for a single polling location.  On average in all precincts, about 8-12.  They could see some Berekeley precincts getting higher numbers as we move forward.  That would equal 4.6 voters provisional voters per hour in the highest precinct.   
 
Okay, thanks for the numbers.  I have polled some folks in the office, and it appears that the consensus is that anyone capable of entering a two digit code is equally capable of entering a six digit code.  Basically, either the poll workers are capable of using a telephone and hand calculator, or they're not.  Hopefully most of them are.
 
The codes themselves are self-checking.  If the operator mis-enters the code printed on the affidavit, then the spyrus unit will reject it.  With a six digits, there are 1,000,000 possible codes they can enter.  9,999,000 of these codes will be invalid.  I'll post a list of codes once we have the coding scheme worked out.  Of course the valid codes would be pre-printed on the pad of affidavits.
They are OK with indexing by number.  But the indexing needs to be thought out.  If we pre-printed numbers on an affidavit, then how would the ballot station automatically add the correct affidavit prefixes so that the poll worker would only have to type in the last 2 digits?  Would we have to pre-print affidavits for each precinct specific to each election??  I don't think that works.  I'm back to barcoding or something. 

No, you just need a standard pad.  Each polling station puts their set of affidavits with the ballots into a big envelope (or whatever) with a vote center label.  On the back end back at election central, the provisional ballots are processed by vote center.  GEMS knows what vote center the ballots came from, and can use that information plus the standard code to find the exact ballot and voter.
I am still hoping that we could manufacture a device that could read a bar code on the application and burn the card vs. trying to have the pollworker use the smaller Spyrus keypad button.   
 
You bring up two issues here. 
 
First is the barcode concept.  The coding scheme makes it very difficult (basically impossible) for poll workers to screw up entering in the number.  I have to have faith that poll workers are capable of operating a hand calculator, or we are probably going to be in a good deal more trouble on other fronts with the touch screen.  We could in theory hook a bar code reader to the spyrus unit, but then you are going to have to power it.  You couldn't wear the reader around your neck, like its designed.  Basically you add an unwieldy piece of hardware to the equation for the exceptional voter.  It doesn't make sense.
 
Your second point is that the keys are small.  Fine.  They're small.  They're also cute.  That they are really too small to be used by the same gene pool that were once charged with hand tallying ballot totals I find difficult to buy.
 
The very idea that we could go out and build a better unit than a company dedicated to building card readers is staggering.  The only serious argument I can see here is that the buttons are too small, and lets face it we are not going to go out and start manufacturing cards readers because the buttons are too small.  We only have so many personnell that are good with silicone glue as it is.
 I guess we'll know more after Texas and Piedmont/Oakland regarding use of the Spyrus.
 
I suspect how well Piedmont/Oakland goes will depend entirely on how good a job you do on selling them on the concept.  If you do not think it will work, I have no doubt that your customer will concur.  In this case, I suggest we start training Alameda poll workers on booting PCs early.
 
Ken