Election UX: Spoiled Ballots

When people vote in elections, they typically do so by casting ballots. These ballots come in many forms, one type being a paper ballot on which the voter indicates each choice by drawing a line between two halves of an arrow. This type of ballot, which is used in various locations within the USA, typically is processed with an optical ballot scanner.

Connect-the-Arrow Voting Instructions

Instructions of this type are often provided on the ballot itself. In addition, there may be
voter information posters and instructions sheets providing examples and illustrations of
how to avoid and correct marking errors.

Without proper instruction, voters may not know how to mark such ballots correctly. For example, a voter may draw an "X" between the two arrow halves, draw a circle around one or both halves, connect the two halves with multiple lines, or unknowingly use the wrong type of writing implement. Alternatively, a voter may know how to connect the arrows correctly, but may attempt to correct a mistaken selection by crossing out, scribbling over, or making notations. When errors of this type prevent votes from being counted, the ballot is described as "spoiled."

Even when a voter understands very well how a ballot should be marked, there still can be mistakes and accidents that spoil a ballot. For example, a ballot may become spoiled by stray marks, smudges, stains, loss of control of a pen, or accidental use of pen that has the wrong color of ink.

When a voter spoils a ballot without recognizing that or without being able to correct that, the integrity of the election is reduced. When the voter recognizes a spoiled ballot and is able to take corrective action, this still may degrade the election experience for that voter or for other voters.

While election integrity is the higher priority, the quality of the election experience is important as well. Each of the use cases described above provides an opportunity to consider the election process and ways in which it might be improved.

Mining Asteroids

Rare Platinum Nugget
U.S. Geological Survey/photo by Chip Clark, Smithsonian

If you drive a car, chances are that it uses a catalytic converter to remove toxic compounds and reduce pollution from fuel exhaust. In order to do this, the catalytic converter uses platinum or its cousin palladium, often in combination with other precious metals such as rhodium and gold.

The metals in the platinum group (platinum, palladium, rhodium, ruthenium, iridium and osmium) have important industrial, medical and scientific applications. They are expensive, however, because their supply in the earth's crust is limited.

In contrast, some asteroids, known as M-type (M for metal), contain rich quantities of platinum and other precious metals, including gold, distributed throughout.  If such asteroids could be reached, mining those asteroids could produce a valuable payload for a returning spacecraft. Before ruling this out as economically infeasible, consider this:

Our solar system contains millions of asteroids. Most of these orbit the sun in a belt between Mars and Jupiter, but there are also near-Earth asteroids (NEAs) with orbital paths either near to or crossing the orbital path of our planet. Fifty years ago, there were fewer than 70 known NEAs. As of February 1, 2014, the IAU Minor Planet Center has cataloged over 10,000 near-Earth objects that are at least 1 meter in diameter. If one or more M-type asteroids could be captured and brought into orbit around the moon, it could be reached for mining by relatively short-haul space travel.

As tantalizing as this prospect may be, asteroids contain another substance that may prove to be even more valuable than platinum or gold. That substance is water, the stuff of life comprising 50-65% of the human body. We cannot exist for long without water. This means that, to exist for long periods in space, humans must have a source of water in space.

Water - The Stuff of Life
Photo by Tim McCabe, USDA NRCS

This is why the first asteroids to be mined are most likely to be C-type (C for carbonaceous) rather than M-type. With a water content ranging up to 20% (though ~10% appears to be more common), these asteroids could provide water to support moon bases or space platforms that would make further exploration and commercial uses of space viable. In addition, volatile compounds found in such asteroids could be used to create fuel, as could hydrogen and oxygen derived from water. This, together with solar energy, could make possible the use of materials found in asteroids to build structures in places beyond our earth. Carbon and organic materials found in asteroids might even be used to grow food. Habitats in space could become nearly or completely self-sustaining.

The future is hard to predict, but baring catastrophic setbacks (perhaps of our own making), human exploration and travel beyond the moon seems inevitable. Asteroid mining may help to pave the way.

Catching Asteroids

Concept Image of Asteroid Capture

Image Credit: NASA

The goal of the NASA Asteroid Redirect Mission (ARM) is to capture a small asteroid and redirect it into an orbit around the moon, where it may become a valuable resource for scientific research, future manned space travel to Mars, and commercial asteroid mining.

The asteroid catching spacecraft to be used for this mission will be powered by advanced solar-electric propulsion, using ion thrusters designed and built at the NASA Glenn Research Center in Cleveland, Ohio.

Ion propulsion works by zapping an inert gas, such as xenon, with an electrical charge to produce fast-flowing exhaust gas.  Providing low thrust over long periods of time, ion propulsion can gradually produce a spacecraft speed of over 200,000 mhp, while using ten times less fuel than traditional rocket engines. This type of propulsion, which has already been used in smaller version by 200 or so commercial satellites since its invention at the Glenn center, makes longer space voyages more feasible.

Current plans call for the robotic capture and redirect spacecraft (which has not yet been named) to be launched in 2017. Various designs for the asteroid capture mechanism are still under consideration. One design concept calls for the use of a large inflatable bag to surround a small free-floating asteroid. Another concept envisions using robotic arms to remove a piece of a larger asteroid.

Meanwhile, work is underway to identify and characterize near-earth asteroids that might be candidates for capture. This effort also will increase our knowledge of asteroid characteristics and help to identify near-earth asteroids that may pose a collision threat, as described in Hunting Asteroids.

Hunting Asteroids

Close-up of Eros taken by NASA's Spitzer Space Telescope
Image Credit: NASA/JHUAPL

In our solar system, there many small rocky objects orbiting around the sun. When these are small, they are referred to as meteoroids; when larger than about 1 meter (3.3 yards), they are referred to as asteroids. There are millions of asteroids in our solar system, and some them -- approximately 89,000 discovered to date -- have trajectories that come near our Earth.

 In June of 2013, NASA announced a project to engage scientists and citizens in helping to "find all asteroid threats to human populations and know what to do about them" through a series of challenges, prizes, and crowd-sourcing activities. While the threat of an asteroid may be low on our list of daily worries, a direct collision is something we'd best avoid. As some may recall, in February of 2103, approximately 1500 people were injured and many thousands of buildings were damaged when a small asteroid exploded over Chelybinsk, Russia. So keeping tabs on near-Earth asteroids seems like a good idea, and one purpose of the Asteroid Grand Challenge is to find better ways of doing that.

 Offered by the NASA Tournament Lab via TopCoder, the first in the series of planned challenges, Asteroid Data Hunter, is now underway. This challenge ask participants to "develop significantly improved algorithms to identify asteroids in images captured by ground-based telescopes," specifying that the "winning solution must increase the detection sensitivity, minimize the number of false positives, ignore imperfections in the data, and run effectively on all computer systems."

Meanwhile, NASA has 50-100 volunteer amateur astronomers from around the globe helping to track and characterize asteroids.

Today at NASA's Asteroid Initiative Opportunities Forum, a group of students (shown above with their teacher) from Dillard Drive middle school described how they're participating in the International Astronomical Search Collaboration by hunting for asteroids. They are the only middle school worldwide to be participating in this activity, and today they reported on their accomplishments so far. You can learn more about this by viewing "Dillar Drive students search for asteroids" on YouTube.

As part of the Asteroid Grand Challenge, NASA is looking for ways to increase the number of amateur astronomers observing asteroids.

In addition to hunting asteroids, NASA is exploring ideas for deflecting and redirecting them. Such techniques might be exploited for research and asteroid mining, as well as for protecting our planet from damaging collisions.

Lava Hits the Beach

After waiting 5 years for just the right conditions, photographers CJ Kale and Nick Selway made this fascinating video of their adventure photographing lava flowing into the sea at the Kalapana ocean entry about 30 miles south of Hilo, Hawaii.




When I visit the Big Island, I plan to stop by the Lava Light Galleries in Kailua-Kona. You may want to check out their online SmugMug galleries to see photos by CJ Kale and photos by Nick Selway. You can learn more about these excellent and daring nature photographers from the Lava Light Galleries Blog.

A Good UI Design Implemented Badly

Screen shot of Photo Tool from travel.state.gov

The U. S. Department of State maintains a website (travel.state.gov) that provides information about passports and international travel for U. S. citizens. If you go to the page for Passport Photo Requirements you'll find a link to a Photo Tool that can be used to crop and size a photo to meet passport application requirements. Using this page, I was able to quickly and easily create a jpeg file sized to make the 2" x 2" print required for a passport application. After using this nifty tool, I wanted to give the web page kudos as an example of good user interface design. Then I realized there are a few problems.

This nice UI design concept has been implemented using Adobe Flash. That means the Photo Tool won't work on most tablets and smart phones -- devices that are likely to contain images that might be suitable for use as passport photos. In order to be successful, a website like this needs to keep the user's needs in mind at every stage.  If it's likely that people will want to use a website on a device that doesn't support Flash, then that should be taken into account from initial design to final implementation.

The Photo Tool also provides an example of how a good UI design concept can be undermined by incomplete implementation. If you look at the screenshot at the beginning of this article, you'll see that there's a help button on the left. This expands when clicked to display a scrollable message, as shown below.

Pop-Out Message Pane

The help button is easy to find and use. However, the help message itself seems to have space for some illustrations that have gone missing. When scrolling through the entire help message (shown to the left), you can see that the UI designer most likely intended to show examples of (a) head sized correctly, (b) head too small, and (c) head too large.

Illustrations here would indeed be helpful, especially if they included graphical symbols to further convey "correct" vs. "incorrect." It may be that the illustrations were not included due to an oversight, or it may be that they're not being displayed due to some technical problem. In either case, this is the kind of problem that QA testing should detect. However, in order to catch such problems, the QA testing needs to simulate likely end-user behavior, and all parts of the design, including help messages, must be complete before final testing.

Even so, a website that's user friendly when it's launched doesn't always stay that way. As technology changes, what worked last year may not work this year. As users adapt to new technologies, their experiences and expectations change. To be successful, then, as website like this also needs to be monitored. Good analytics can often reveal problems or new trends, signaling a need for action. End users also can help you know when changes are needed -- if you let them.

Although the webpage for the Photo Tool has no links for user feedback, it's parent page (Passport Photo Examples) has "Contact Us" links at the top and the bottom of the page. The link at the top leads to a page providing categories for contact, and the link for "Website Problem or Suggestion" leads to a well-designed feedback form. However, the other "Contact Us" link takes the user down a different path from which there's no obvious way to provide feedback on the website.

(Post edited on March 8, 2014 to improve clarity.)

Addendum - March 8, 2014

Website development, like all software development, requires tradeoff decisions. Given a choice between having the Photo Tool work only on Flash capable devices vs. not having it at all, I'd obviously choose the former. I liked using the Photo Tool and I was glad to have it available. Indeed, I originally planned to write an entirely positive post citing the Photo Tool as an example of good UI design. Taking a closer look, however, I was reminded of an important point:  Usability involves more than good UI design.

Searching for Extraterrestrial Intelligence in Space and Time

Voyager 1 - Image credit: NASA/JPL-Caltech

It's estimated that there are about 30 billion trillion stars in the Universe. Recent scientific discoveries suggest that many of those stars may have planets. Indeed, current evidence suggests that there may be billions of planets similar to our Earth in their ability to support life. Is it therefore likely that we, the human inhabitants of Earth, will someday discover or make contact with extraterrestrial intelligence (ETI)? Perhaps. But in considering that question, we have to take into account the magnitude of both time and space.

Scientists currently believe the age of our Universe to be about 13.8 billions years old. Our Sun is thought to be about 5 billion years old. There's evidence that simple-celled life first appeared on earth about 3.6 billion years ago. Fossil remains suggest that homo habilis emerged about 2.5 million years ago, and homo sapiens has been around perhaps 200,000 years -- though estimates vary, as do theories about when modern human behavior emerged.

To give you an idea of how short human existence has been, if we consider the age of the Universe to be one year that began on January 1, then our sun would have appeared around August 20 and our earth around September 3.  Life in the form of simple cells would have been present near the end of September, and animals in the middle of December. Mammals would have come around December 26 and primates around December 30. The first hominids would have shown up around noon on December 31, and the first members of the genus homo would have arrived around 10:30pm. Homo sapiens would have appeared anatomically sometime sometime during the last 10 minutes of the year, begun to exhibit modern behavior perhaps in the last 2-3 minutes, and invented writing sometime during the last 15 seconds. Within this metaphorical cosmic year, the time of homo sapiens will that of one dance if we survive and thrive for another 50,000 years, and that of a thunderstorm if we last another 500,000.

Biological processes tend to be self-limiting. On earth, species have come and gone. If this is our fate and the fate of all intelligent life in the Universe, the chances of encounter depend greatly on how long it takes intelligent life to develop and how long it can survive. Let us suppose there are 100 trillion planets that will, at some point in their existence, have intelligent life capable of reaching out. If that were to happen at random for the period of one dance during my metaphorical cosmic year, we would have several hundred thousand potential partners. If the time of reaching out is more like the metaphorical thunderstorm, the number of potential partners increases to over half a billion. This may seem like good odds, but we also must take into account the vastness of space.

Consider the distance between galaxies. The Andromeda galaxy, a near neighbor of our own Milky Way, is about 163,000 light years away. If an observer within Andromeda were to have a telescope powerful enough to see the surface of the earth in detail, looking today they would be seeing the earth as it was ~163,000 years ago. Thus, they might see hominids (including Neanderthals) making tools of bone and stone, living in shelters, wearing clothing and controlling fire. That observer might conclude that there was intelligent life on earth, and might also decide to attempt communication in hope that by the time the signal arrived, the intended recipients would have the technology to detect it. In 163,000 years or so from today, that signal would arrive.

Given the time-distance involved, unless we learn how to communicate faster than the speed of light, it seems unlikely that we'll be exchanging signals any time soon with intelligent life outside our own galaxy. However, within our own Milky Way, which spans at least 100,000 light years in its longest direction, the Kepler space telescope has found over 1700 planets that are within 3,000 light years from earth. That's just since 2009, and our ability to find planets has recently gotten better. We now have reason to believe that there are many planets out there orbiting stars whose distance in light years is comparable to the 5,000 or so years since the earliest known evidence of writing.

Now, as a hypothetical exercise, let's suppose that the ~600 million stars within that distance have an average of one planet each that's in the habitable zone, and that complex life develops on 10% of those. If intelligent life is but a dance occurring randomly within the cosmic year, we would have 100 or so potential partners. If it's a thunderstorm, that number expands to over 3,000. If, on the other hand, it takes about the same amount of time for intelligent life to emerge on each planet, our chances of detecting it would be much greater. In addition, there's the chance of finding evidence of an intelligent civilization that once was and is no more.

Even so, what these back-of-the-envelope calculations suggest to me is that the search for ETI is an act of optimism, grounded in the hope that intelligence confers the ability to plant a continuing harvest for posterity rather depleting the soil or sowing the seeds of its own destruction.

A Small Neighborhood in Fornax

My last post was about an image I created using FITS data I got from SkyView. Here's an image from NASA showing a larger view of the same region in space. This is a small area in the celestial sphere, spanning only one minute of arc. It's located in the constellation Fornax, which occupies an area of 398 square degrees. The name Fornax comes from the French astronomer Nicolas Louis de Lacaille, who first formed this constellation and called it Fornax Chimiae ("chemical furnace").

The constellation Fornax can be seen from the southern sky, but the stars and galaxies shown in the image above cannot be seen with the naked eye. They are in what's known as a "dark field," a region where there is not much background radiation. By surveying such regions with the Hubble Space Telescope, astronomers are able to detect and study faint galaxies in the early universe to learn more about how galaxies evolve.

FITS Liberator Image from 03h32m29.3s, -27d44m10.0s (J2000)

This is an image I created today using the FITS Liberator with data obtained from SkyView. What you're seeing is an area of deep space that was surveyed by the Hubble Space Telescope Advanced Camera for Surveys (HST ACS). I got the FITS data by giving SkyView the coordinates 03h32m29.3s, -27d44m10.0s (J2000 co-ordinate system) and specifying regions .02 degrees in size.

To get a color image, I downloaded three sets of FITS data, each covering the same area of space, but with different filters for wavelength. Using the FITS Liberator, I converted each data set into a TIFF image. I then loaded the TIFF images in to Adobe Photoshop to make a stack of three layers and gave a different color (i.e. red, green, blue) to each layer. After making some curves adjustments, I got the image you see above.

Making this image involved a series of subjective judgements, first in managing the dynamic ranges of images with the FITS Liberator, and then in tweaking curves in Photoshop. So the image you see should be considered an artistic rendering of astronomical data rather than a "snapshot" of outer space. I did, however, try to keep the colors basically "natural" by assigning blue to the shortest frequency filter, red to the highest frequency, and green to the one in the middle.

When looking at my image, you may be wondering if there's a way to describe the part of space it shows other than by the coordinates I've provided. I was wondering that too, at first!  I've never taken a class in astronomy, and I understand less than half of what I read on the subject. However, I'm fascinated by astrophotography. So my approach to learning about astronomy has been to find interesting images and then try to understand what they show. If you're an astronomy buff, you may know what my image shows. Otherwise, you're going to have to wait for me to post more information.

The Asymmetric Internet

As the Internet was opened to commercial use in the 1990s, access became available to people outside of academic and research organizations. Various companies -- large and small, local and national -- began to offer dial-up access to the Internet for people who wanted to exchange information by such methods as email, newsgroups, file sharing and online chat. This information sharing typically involved a mix of "upstream" and "downstream" data transfer between users and ISPs. For example, when user A sent user B an email message, information went upstream to A's ISP (X), across the Internet to B's ISP (Y), and then downstream to B. When user B sent an email reply, information was sent in the opposite direction.

The dial-up connections that people used to share information across the Internet in this way were symmetric, in that they were able to transmit data between the user and their ISP both upstream and downstream with equal facility. The same was true for the higher-speed telecom connections between IPSs. In the early days of the Internet, these symmetric connections were intended to facilitate long-distance collaborations in which the various participants all had information and ideas to share.

As the Internet grew and evolved, people began to feel constrained by the data transfer speeds of dial-up connections. As DSL and cable connections became available, they were rapidly adopted by those seeking greater bandwidth. These new technologies were deployed, in most cases, to provide higher data rates downstream than upstream, making them different from the telecom connections that had previously been used for Internet connections.

Today we have an Internet over which a small number of "content providers" send large amounts of data downstream to end users. Is this a natural evolution to which technology has adapted, or has the technology itself produced this result?

I understand that certain kinds of content are easier to consume than to produce, leading to a natural asymmetry in the proportion of writers to readers, performers to audience, etc. However, the skills and talents required to produce good content are, I believe, distributed around the globe. Furthermore, the kind of data that can be carried by the Internet can be used for many purposes. So my questions are these: If we had an Internet that gave us all a high-speed connection both upstream and downstream, would the way we use the Internet change? Would the Internet be used for two-way communication and collaboration more than it is today? Would more good writing, good photography, good music, good video, good ideas be available to us? Would innovators be stimulated to create new technologies and new applications?

Why do we have an asymmetric Internet? Is what we get based on what we want, or is what we seem to want based on what we get?

From Citizen to Consumer

The chart above, which I produced using the Google Ngram Viewer, shows how use of the term "citizen" has declined, while use of the term "consumer" has increased.

In the context of discussions about public policy, I prefer being referred to as a "citizen" rather than a "consumer." To me, the word "citizen" connotes an agent who acts and makes decisions within a matrix of rights, responsibilities and ethical values that's greater in scope than the economic decision matrix implied by the word "consumer." This is not to claim that people make consumer choices based on economic calculus alone, but rather to suggest that deviations from this calculus tend to be seen as just that -- deviations -- when the agents making choices are thought of as "consumers" rather than as "citizens."

However, the nuanced meaning of words and the cloud of associations they evoke change over time. In her book A Consumer's Republic: The Politics of Mass Consumption in Postwar America, Lizbeth Cohen discusses how the identification of American citizens as "consumers" shifted during the 20th century from a grounding in ideas of civic responsibility and democratic values to an insular emphasis on personal benefit. Such shifts in meaning can not only reflect change but also create it, as words have the ability to shape perceptions. (Cohen also has written a short article summarizing the main points of her book.)

In their book Citizens or Consumers? What the Media Tell Us about Political Participation, Justin Lewis, Sanna Inthorn and Karin Walh-Jorgenson describe the results of their research based on empirical data from 5,658 news stories that were published in Britain and in the United States. Stuart Allen, series editor, describes the book as follows:

"This book enters squarely into the current controversy about the declining number of people engaged in the conventional political process. Based on the first comprehensive study of the way citizens are represented in the news, it provides powerful evidence that while the news media may not be responsible for this decline, they are doing little to help remedy it. Although many people do have clear ideas about politics and matters of public policy, the authors argue, the news media are much more likely to present citizens as passive, incoherent or or disengaged. Their research into news reporting suggests that the idea of the citizen has been more or less eclipsed by the figure of the consumer. So while we hear a great deal about the 'consumer confidence' on which economies depend, we hear very little about the 'citizen confidence' on which democracies depend."

Perhaps it's time to reverse the trend and bring the word "citizen" back into our daily discourse.