Voting White Paper

www.benchmarkrs.com | info@benchmarkrs.com

PRACTICAL SOLUTIONS FOR A COMPLEX WORLD

Voting is as simple as 1,2,4...
The Human Factors of Voting
Eric F. Shaver, Ph.D. and Curt C. Braun, Ph.D. October 2008

The right to vote is one of the fundamental liberties granted to American citizens
by our Constitution. It is our duty to employ this right in order to maintain a
free society. Ideally, all citizens who chose to exercise their right to vote should
be afforded the opportunity of doing so. In theory, the process of casting a
ballot appears simple. Unfortunately, even in the 21st century, the process of
voting and vote counting is so varied and complex that confidence in the system
wanes. Differences occur in voter registration, absentee and early voting, ballot
design, the type of equipment used to cast votes, vote counting, and certification
of the vote. What is needed is a simplified process that minimizes voter error
(i.e., overvotes, undervotes, etc.) and guarantees that cast votes will count. When
these requirements aren’t achieved, the system breaks down and our democracy
is put at risk.

Voting Issues in the 21st Century

Voting issues are not new. Since the first election, officials have had to deal with
issues such as who has the right to vote, election tampering, registration
inaccuracies, and vote tabulation errors. In these situations every form of voting,
from paper to technology, have failed to withstand scrutiny. Although voting
issues are as old as voting itself, problems surrounding the act of casting a vote
came to the forefront in the 2000 presidential election.

2000 Presidential Election

In the 2000 presidential election, a poorly designed ballot used in Palm Beach
County, Florida contributed to a disproportionate number of votes being
unintentionally cast for Reform candidate Pat Buchanan, instead of the intended
recipient, Democratic candidate Al Gore. The problem with the butterfly ballot
stemmed from its double-column design. The two main party candidates,
George W. Bush and Al Gore, were located in the first and second positions on
the left-hand column, but their corresponding punch holes were in the first and
third positions, while the second punch hole was reserved for Pat Buchanan who
was located in the first position on the right-hand column. Thus, people who

1 PO BOX 9088 | Moscow, ID 83843 | tel: 877.641.4468| fax: 208.882.2541

intended to vote for Al Gore by punching the second hole instead voted for Pat
Buchanan.

The overall state vote yielded a difference of 1,784 votes in favor of Bush,
triggering an automatic recount. After the recount the difference was decreased
to 537. During the recount process, a further ballot design flaw was revealed
with the existence of hanging, swinging, tri-, and dimpled chad (paper debris left
attached to a ballot of a punched hole), and votes were not counted. Moreover,
19,120 ballots were considered “spoiled,” and thus not counted, because voters,
in an attempt to correct their “mistake,” punched a second hole after realizing
they had voted for Pat Buchannan and not Al Gore (Vicente, 2004). Ultimately,
due to unintentionally miscast and uncounted votes, it was the US Supreme
Court, and not the voters, who determined the outcome of the election in favor
of George W. Bush.

Help America Vote Act of 2002

The 2000 presidential election problems highlighted the need to reform the
voting process, a goal of the Help America Vote Act of 2002 (P.L. 107-252).
This act:

• Provided states the funds to replace punch card machines with electronic
voting machines
• Established the independent, bipartisan Election Assistance Commission
(EAC)
• Developed minimum election administration standards for states to
follow

Nationally, election officials reviewed their voting procedures and many moved
to modernize by replacing antiquated voting equipment (e.g., paper ballots,
mechanical lever machines, and punch cards) with more contemporary voting
methods (e.g., optical ballot scanning and direct recording electronics). For
many, technology was the solution to their problems. Unfortunately, instead of
fixing existing problems, this new technology replaced old problems with new
ones.

2004 Presidential Election

During the 2004 election, these new problems became apparent. New electronic
voting systems were plagued by a variety of issues including reliability, security,
auditability, and inadequately trained poll workers (GAO, 2005; Kohno, et al.,
2004; Lovgren, 2004; Selker, 2004). Concerns about these issues were borne out
during a 2006 election in Florida.


2006 Florida 13th Congressional District

The 2006 election for the 13th Congressional District of Florida consisted of a
tight race between Republican candidate Vern Buchanan and Democratic
candidate Christine Jennings, with the former winning by 369 votes. Voters used
the ES&S iVotronic direct recording electronic voting machines to record their
votes. Given the close finish, a mandatory recount was performed which
consisted of re-querying the machines, and Mr. Buchanan was again declared the
winner. Mrs. Jennings challenged the results, among other reasons, due to an
alleged voting machine error. Her protest centered on the fact that 18,000
people (15% of the voters), who cast votes for other offices, did not record a
vote for either her or Mr. Buchanan - referred to as an “undervote.” It was
suggested that the design of the electronic touch screen ballot contributed to the
undervote because it inconsistently used a blue banner to highlight the races.
Jennings argued that voters may not have noticed the 13th Congressional District
race on the ballot because the banner was missing. Investigations by the
Government Accountability Office have been inconclusive (GAO, 2007).

A common denominator in many voting issues, particularly those that involve
the actual act of casting a vote, revolves around an inadequate understanding of
the qualities, characteristics and differences among voters, including how they
interact with voting systems. The discipline of human factors can shed valuable
insight that can promote the reliability and effectiveness of today’s voting
systems.

What is Human Factors and Why is it Important to Voting System
Issues?

Human factors is a unique scientific discipline that systematically applies the
knowledge of human abilities and limitations to the design of systems with the
goal of optimizing the interaction between people and other system elements to
enhance safety, performance, and satisfaction. In simpler terms, human factors
focuses on designing the world to better accommodate people. In the case of
voting, it would focus on making the process of voting compatible with voters.

Voting, although simple by definition, is complex when put in practice. Voting
involves voters, poll workers, election officials, voting equipment, ballots,
booths, facilities, registration, absentee and early voting, counting, recounts,
certifications and all of the individual processes that make recounts and
certifications possible. All of these tasks and sub-processes are interdependent
so that changes in one aspect of the process are likely to affect other process
components. Let’s look at the large components of the voting process: people;
voting technology; voting task; and voting environment.


People

Three groups of individuals are vital to the voting process – voters, poll workers,
and election officials. The diversity of the voting public, while great from a
multicultural perspective, adds significant challenges to designing voting systems
that are sufficiently accommodating. These differences include first-time voters,
older adults, low-literacy, limited English skills, people with disabilities, and
novice computer users. Thus, it is important to provide the necessary
accommodations for each of these groups to foster successful voter
participation. This would include providing voter education materials in multiple
formats (e.g., text, audio, video, etc.), via different avenues (e.g., mail, email,
Internet, etc.), and in multiple languages. Also, voters should have the
opportunity to see their specific ballot prior to voting so they can familiarize
themselves with its format and content.

Poll workers are important to the voting process in several ways. They ensure
voting progresses smoothly, aid voters who need assistance, and ensure the
polling place is safe and secure. Despite their key role, many poll workers are
unable to adequately explain the use of voting technology, are unable to properly
instruct voters in its use, and cannot problem solve issues onsite. Interestingly,
many of the problems stem from poorly designed voting technologies and not a
lack of training. But, in the absence of properly designed and tested voting
technology, poll workers need to be provided with sufficient training so they can
successfully assist voters who need it.

Election officials play a vital role in making sure the voting process performs
smoothly for voters. Their importance comes to the forefront when challenges
like the 2000 and 2004 Presidential elections occur. Within state government,
the Secretary of State oversees the electoral process for their state, which
includes putting out voter guides, determining how funds are spent on the
procurement of voting equipment, how electoral reform should progress, and
certifying candidates and election results. Given the lack of uniformity of the
voting process across the US, the importance of the decisions they make cannot
be overstated. Thus, it is imperative that only voting technology that conforms
to basic human factors principles be purchased and implemented.

Voting Technology

Voting Methods

Currently, there are five different methods for voting in the US - paper ballots,
mechanical lever machines, punch cards, optical scan, and direct recording
electronic. Three of the methods make use of computer technology (punch
cards, optical scan, and direct recording electronic) and three have some form of
a paper trail (paper ballots, punch cards, and optical scan). With respect to


Idaho, three methods are used – paper, punch card, and optical scan (arrow &
oval).

Paper ballots have been used since our independence, with the secret ballot being
introduced at the end of the 19th century. This method has voters mark a box
placed next to their voting choice and completed ballots are placed in a sealed
ballot box.

Mechnical lever machines were first introduced in the late 19th century and were the
most common method of voting through the mid 1980s. This method has
voters activate a lever next to their voting choice and the machine records it on a
vote counter that is opened and tallied by election officials at the end of the
voting day.

Punch cards were first implemented in the early 1960s and replaced the mechanical
lever machine as the most frequently used method until the 2000 elections. This
method was the first to use computer technology by incorporating a computer-
readable card that required voters to use a stylus to punch out a box next to their
desired voting choices. The votes are tallied by feeding the card into a tabulation
machine that electronically reads the punched holes.

Optical scan systems were first used in the 1960s and currently are the most
common method of voting. Voters fill in either an arrow or oval next to their
voting choice on a computer-readable paper ballot and the completed ballot is
fed into a tabulation machine that reads and tallies the marks.

Direct recording electronic (DRE) was introduced in the mid-1970s and currently is
the second most used method. This method has voters either push a button or
touch the screen next to their voting choice. The completed ballot is submitted
by pressing the vote button, which stores their votes on the computer’s memory
chip

Each of the five methods has its own set of potential issues:

• Paper Ballots - stuffed, lost or stolen ballots and/or boxes
• Mechanical Lever Machines - no voter-verifiable paper audit trail
• Punch Cards - hanging, swinging, tri-, and dimpled chad
• Optical Scan - computer malfunctions (e.g., misreads, crashes, etc.);
Inability to recognize ink used to mark ballot
• Direct Recording Electronic - security; poll workers need greater levels of
technical experience; no voter-verifiable paper audit trail

Ballots

Depending on the voting method, ballots come in one of two basic forms -
paper or electronic. These, in turn, can present the entire ballot on one
page/screen or separated into many individual pages/screens. On top of these


differences, ballot choices and design often vary between districts, states, and
elections. Moreover, besides just listing candidates for federal, state, and local
offices, ballots often include bond issues, propositions, and constitutional
amendments – many of which are written in confusing language. Each of these
factors contributed to the complexity of the ballot, thus increasing the likelihood
of voting issues such as overvotes, undervotes, and uncounted ballots to occur.

A number of authors (Herrnson, et al., 2008; Kimball & Kropf, 2005; Niemi &
Herrnson, 2003) have identified issues that should be considered when designing
ballots, including:

• Readability - upper & lowercase san-serif type; don’t use all caps;
minimize number of fonts; left justify text, etc.
• Comprehensibility - clear instructions; simple language; consistent use of
candidate names; etc.
• Layout - order of candidate’s names; candidates for the same office listed
in one column or page; minimize clutter; avoid mirror images; etc.
• Features - appropriate use of color, highlighting, shading, bolding, etc.

Voting Booths

Voting booths allow voters to cast their ballots in private, while improving
voting accuracy by minimizing distractions. They take many forms from curtain-
partitioned stalls to individual stations divided by screens. Concerns that have
been raised include the instability of fold-up stands used by some voting
methods (Herrnson, et al., 2008). Moreover, the height of the voting booth can
make it challenging for voters with height extremes (i.e., short or tall) to
comfortably complete the voting process.

Voting Task

There are several factors that can negatively impact voting and need to be
considered when designing voting systems. These include increased mental
workload when using poorly design voting technology, increased time pressure
due to long voting lines, and distraction from a noisy voting environment. Each
of these factors can contribute to a situation where errors might invalidate a vote
for a given office – or the entire ballot. Also, when the voting task becomes too
challenging, voters are forced to enlist the assistance of poll workers, thus
suffering a loss of voting privacy.

Voting Environment

Typically, voting is held in facilities (e.g., schools, churches, fire stations, etc.)
that are less than optimal environments for tasks requiring concentration. Added
to this, are long voting lines, high ambient noise levels, having to deal with
pollsters when entering and exiting the voting facility, and the potential for
accessibility issues. Each of these issues needs to be considered and addressed at


every polling place to ensure an environment that minimizes the likelihood for
voter error.

How Can Human Factors Minimize Future Voting System Issues?

Heuristic evaluations and usability testing are two human factors methods that
can be used to analyze current voting systems, along with voting technology
prior to procurement and implementation. These methods are directly applicable
to voting systems and can be performed either by election officials or with the
assistance of a human factors consultant. Ideally, both techniques will be used,
given that they each have their own strengths and weaknesses.

Heuristic Evaluations

Heuristic evaluations consist of expert evaluators using a list of human factors
principles (heuristics) to independently identify usability issues with a given
device. The evaluators should be individuals who are familiar with the voting
system, potential conditions under which voting will occur, and have an
understanding of human factors. Heuristic evaluations consist of three steps:
comparing the heuristic list against the voting technology, identifying heuristic
violations, and determining the severity of each violation. For example, one set
of heuristics used to evaluate voting equipment (Herrnson, et al., 2008, p. 37),
includes:

• Simple and natural language - Messages should not contain information
that is irrelevant or rarely needed.
• Speak the user’s language - Dialogues should be expressed clearly in
words, phrases, and concepts familiar to the user rather than in system-
oriented terms.
• Minimize the user’s memory load - The user should not have to
remember information from one part of the interface to another.
• Consistency - Users should not have to wonder whether different words,
situations, or actions mean the same thing.
• Feedback - The system should always keep users informed about what is
going on through appropriate feedback within reasonable time.
• Reversal of actions - As much as possible, actions should be easily
reversible.
• Clearly marked exists - Users should be able to leave the system at will
and not need assistance to do so.
• Good error messages - All errors should be avoided if possible, but any
error messages should be expressed in plain language (no codes) precisely
indicating the problem and constructively suggesting a solution.
• Help - Even though it is better if the system can be used without
documentation, it may be necessary to provide help. Any such
information should be easy to navigate and should be focused on the
user’s task.


Heuristic evaluations are used for three reasons. First, they are used to identify
potential usability issues that might arise from the voting technology itself or
after its integration in the current voting system, which in turn, might lead to
voting errors. Second, they are useful when comparing multiple voting devices
by helping to “weed out” voting technology that violates basic human factors
principles, thus saving the time and cost of performing subsequent usability tests.
Finally, the outcome of heuristic evaluations can help guide training efforts by
identifying what needs to be addressed when training poll workers and educating
voters.

Heuristic evaluations have several strengths and weaknesses, including:

Strengths Weaknesses
• Easy to learn and use • Does not discover missing
• Requires minimal training functionality
• Cost effective to implement • Relies on the knowledge and
the experience of the
evaluator

Usability Testing

Usability testing consists of having voters perform the process of voting with
actual voting technology in a realistic setting. The steps required to complete a
usability test include: (1) design the test; (2) performing a dry run and making
modifications to the test design if necessary; (3) recruiting voters; (4) conducting
the test; (5) analyzing the results; and (6) determining the appropriate action to
take based on the test results.

Usability testing can capture objective and subjective outcome performance data.
The former typically includes time to perform individual votes, voting accuracy,
and number and type of errors made by the voters. The latter includes user
feedback regarding strengths and weaknesses of the voting technology and
perceived ease of use.

Usability testing has several strengths and weaknesses, including:

Strengths Weaknesses
• A small number of voters are • Can be time consuming and
needed to identify most costly to perform
problems • Requires testers that are
• Identifies the severity of knowledgeable with the
voting problems method
• Detects voting problems not
identified by heuristic
evaluations


Final Thoughts

Election officials are encouraged to incorporate both heuristic evaluations and
usability testing to identify potential problems with the current voting process
and to determine how best to implement voting technology in the future.
Heuristic evaluations are useful for identifying potential usability issues that
might occur from the voting technology itself or after its integration in the
current voting system. However, heuristic evaluations cannot capture all the
potential usability issues for new voting technology or the level of voter
acceptance and voting accuracy. Thus, it is necessary to also conduct usability
testing with voters in a realistic setting to identify problems that might arise
during the voting process. While both methods have their place, when used in
tandem, they can have the greatest impact and success. Ultimately, both
methods help achieve the overall goal of minimizing voter error by improving
voting system design.

Most of the previously discussed issues from the 2000, 2004, and 2006 elections
could have been prevented by the appropriate implementation of human factors
methods, thus making the voting process compatible with voters. Unfortunately,
failure to do so undermined voter confidence, led to feelings of disenchantment
by those whose votes were not counted, and raised concerns about the integrity
of the voting process. Even more seriously, in the case of the 2000 presidential
election, it removed the power from the hands of the electorate and temporarily
placed it with the judicial branch.

References

GAO (2005, September). Elections: Federal efforts to improve security and reliability of electronic
voting systems are under way, but key activities need to be completed (GAO-05-956).
Washington, DC: United States Government Accountability Office.
GAO (2007, October 2). Elections: Further testing could provide increased but not absolute
assurance that voting systems did not cause undervotes in Florida’s 13th congressional district
(GAO-08-97T). Washington, DC: United States Government Accountability
Office.
Herrnson, P.S., Niemi, R.G., Hanmer, M.J., Bederson, B.B., Conrad, F.C., & Traugott,
M.W. (2008). Voting technology: The not-so-simple act of casting a ballot.
Washington, DC: Brookings Institution Press.
Kimball, D.C., & Kropf, M. (2005). Ballot design and unrecorded votes on paper-based
ballots. Public Opinion Quarterly, 69, 508-529.
Kohno, T., Stubblefield, A., Rubin, A.D., & Wallach, D.S. (2004). Analysis of an
electronic voting system. Proceedings of the 2004 IEEE Symposium on Security and
Privacy, 27-40.
Lovgren, S. (2004). Are electronic voting machines reliable? National Geographic News.
Niemi, R.G., & Herrnson, P.S. (2003). Beyond the butterfly: The complexity of U.S.
ballots. Perspectives on Politics, 1, 317-326.
Selker, T. (2004). Fixing the vote. Scientific American, 291(4), 90-97.
Vicente, K.J. (2004). The human factor. New York, NY: Routledge.


Shaver Braun

Eric F. Shaver, Ph.D. is a senior consultant with Benchmark Research & Safety, Inc.,
based in Boise. Dr. Shaver specializes in human factors & ergonomics, usability, applied
decision making, and safety. Dr. Shaver's work has emphasized achieving a good fit
between people and technology to facilitate their safety, performance, and satisfaction.

Curt C. Braun, Ph.D. is the president, CEO, and founder of Benchmark Research &
Safety, Inc. Dr. Braun has brought psychological and human factors principles to a
variety of industries, including aviation, software development, public administration
and research, and wildland fire management. In each field, Dr. Braun has worked to
identify and shape the psychological and system design factors that promote human
performance.

Benchmark Research & Safety, Inc. specializes in providing consulting and professional
services for a variety of areas including human factors design & usability, product &
occupational safety, and training & education. Our ability to blend technology and
psychology while seamlessly bridging the gap between academia and business is what
sets us apart from other consulting firms. For more information on how our
experienced staff can develop a personalized solution that best fits your needs, please
contact Dr. Shaver at eshaver@benchmarkrs.com or 208-407-2908.


Voting White Paper

More Related Content

Similar to Voting White Paper (17)

More from Benchmark Research & Safety, Inc. (9)

Voting White Paper