HFP Subcommittee Teleconference
Friday, November 9, 2006

Participants: Alice Miller, Allan Eustis, Sharon Laskowski, Tricia Mason, Wendy Havens, Whitney Quesenbery

Agenda:

a) TGDC and EAC updates from Allan Eustis
b) Sharon Laskowski will go over items that will be part of the HFP presentation at the TGDC December meeting. And, the most current VVSG07 HFP. A nicely formatted version of this will be what is distributed to the TGDC for the meeting.
c) Sharon Laskowski will summarize status of current research in more detail.

Administrative Updates:

  • Allan: Reminder to all about reservations for traveling to the TGDC. All but a few have made travel plans. Any questions or concerns should be directed to Lucy Salah.

  • Allan: On Monday 11/13, our TGDC subcommittee chairs will teleconference in to our monthly NIST/EAC meeting . It's a coordination effort to ensure everyone is on the same page for the TGDC meeting.

TGDC Presentation Summary:

  • Sharon sent around the latest VVSG which hasn't been formatted for word.

  • Unless anyone has anything specific about the draft, discussion will be on the specific items suggested for the December meeting.

  • HFP plans to talk about the following items based on the VVSG07 Section 3, Usability and Accessibility draft and related discussions. (Whitney had expressed great concerns about making the document readable, that is expressing things clearly and simply):

    • We added and/or rewrote material for a more readable document. Clarified HAVA requirements and VVSG requirements. There has been discussion about moving this as well up front to the introduction section.

    • What will the usability performance requirements and test methods look like? Open issue. Not sure what the requirements will look like because we're in the middle of the research. Detailed tasks out, will get benchmark for error rate as a whole instead of detailed analysis. Taken out specific function requirements to simplify. Policy issue - how public should the test reports be? Especially the performance benchmarks. Current thinking is Public Report to include all but specific proprietary info. Plans to create test report that is similar to the format of the ISO standard.

    • Clarified requirements for alternative languages in section 3.2.6 because language in VVSG 05 ambiguous - no feedback from EAC, it is needed here.

    • Is the TGDC satisfied with our treatment of marginal marks (new requirement handed over from CRT)? 3.2.2.2 A marginal mark is one in a non-editable interface which is not clearly countable as a vote or as a non-vote - e.g., bubble half filled in. Each vendor must clarify what they consider a marginal mark and what their tabulator can identify.

    • Updated the user interaction timing issues; there is new material. Section 3.2.5.1. We did email TGDC members who were previously interested, so far no response.

    • Consolidated and clarified voter's choice issues (regarding under and over voting). Same process to interested TGDC. System should allow corrections before vote is cast.

    • New material on plain language and usability issues based on general guidelines. Researching specific guidelines.

    • Received from David Baquis a small portion of an ISO standard draft. David has also shown the VVSG work to Access Board for comment.

    • New section 3.2.8.1 that incorporated/consolidated usability for poll workers and moved and clarified maintenance and safety to our section. Whitney would like a statement about the importance that the documentation be complete and usable (effectively) by the poll worker during the operation of an election - make sure we have a statement about who the documentation is for. Sharon will double check that this is going into the user documentation section. (See Sharon's #10 discussion item.) We plan to bring in a documentation expert for assistance.

    • Tried to clarify voter control of color, removed Appendix D because it based on old guidance. Not happy with it, bring in a color expert for advice. Sharon attended a conference in Baltimore about state of the art color realization. Whitney: just make sure expert knows about disabilities.

    • Universal usability. This could be controversial. Long term goal is a universal design that addresses as many people with different capabilities as possible in the general voting section. Example, some machines allow you to adjust font size; this is just not a feature on the "accessible" machines. Is there anything in the accessible systems that should be moved into the general machine guidelines? Think about things that might be a "should" in the general section, and a "shall" in the accessible section.

    • Research is underway for plain language specific to voting--no results to report yet. There's a glitch in the testing. Validation software issues, we've made significant changes to our test ballot.

    • Our usability test protocol and benchmark efforts have not yet addressed the ACC-VS. Work need to be done. If this is not going to make it into VVSG 07, we need a white paper to outline the issues - something that says this is the base and the modifications that need to go into it.

    • Research is underway to develop usability performance benchmarks. How much should we report? We do not have written material on the data collected. We need larger population to benchmark and show repeatability. Next step is starting now. There's more undetected errors out there than be reported. Should this information be discussed at the TGDC meeting since it is a public forum? Satisfaction surveys not useful as a benchmark, because everyone thinks they voted correctly. Not a big difference in timing between DREs and op scans. (Point of reference, this is a public phone call.) Issues to mention some of the other papers, which Whitney can do since she has reviewed them. We can contact the researchers and ask for summaries. We can do a ballpark presentation on the kind of error rates we're seeing. The satisfaction and confidence survey data may play a part when looking at accessible systems. Error rate is the benchmark we want to use.

    • Heads up: We need to start talking in terms of actual things to do in regards to HFP requirements that overlap vis a vis security and usability. Maybe user testing on different security methods proposed. This is research that doesn't exist yet.

Next meeting is Friday, November 17, 2006, at 11:00 a.m. ET.


**************

Link to NIST HAVA Page

Last updated: July 25, 2007
Point of Contact

Privacy policy / security notice / accessibility statement
Disclaimer / FOIA
NIST is an agency of the U.S. Commerce Department