A Timely Article on Voting Machines

With the U.S. election only a week away, I bring you a tale of warning from Finland.

Today, the Ministry of Justice revealed that due to a usability issue, voting was prematurely aborted for 232 voters. The pilot system was in use in three municipalities; this amounts to about 2 per cent of the electoral roll. Seats in the municipal assemblies are often determined by margins of only a couple of votes.

It seems that the system required the voter to insert a smart card to identify the voter, type in their selected candidate number, then press “ok”, check the candidate details on the screen, and then press “ok” again. Some voters did not press “ok” for the second time, but instead removed their smart card from the voting terminal prematurely, causing their ballots not to be cast.

A designer might negate the problems of 2% of users for some systems, but voting machines should be held to to a higher standard. You might not catch 2% with Nielsen’s standard 4-6 users testing the system (or “none” as I suspect was true of the Finnish system.) If you want to judge for yourself, a flash demo of the system is available here. They apparently enter a number code for a candidate.

Input devices were also at the forefront in Finland:

[Added 29th Oct:] There has now been at least one report of touchscreen issues. A voter had repeatedly tried to click on “ok”, but either due to system lag or touchscreen sensitivity problems, it took “minutes” to get the button press registered. If hit by this type of problem, the voters may well have thought that the ballot casting process had completed.

Finding sources for this post was difficult, as most of the sites were in non-translatable Finnish, a language not supported by Google or Babelfish. I would especially like to know the age of the touchs screen user and if s/he had any defining characteristics that might relate to the difficulty with the screen.

2% may not sound like much, unless you’re in the 2%. I suppose the most comforting knowledge is that usability issues probably won’t affect one party more than the other.

Amazon Windowshop

Amazon has a new interface for “window shopping” on its website.  It is strangely compelling and offers something that the website lacked–that quality of just browsing what’s there without having to do too much clicking, searching, etc.  The navigation is simple and intuitive (right/left arrows) and if you stand at a window for a few seconds, a video description starts playing.  Press spacebar to “zoom out” for a more global view.

It feels like Google Maps–the “manipulatability” is satisfying.

Update:  Upon further reflection, the animation probably contributes to visual momentum.  Visual momentum is one reason why, possibly, sliding ipod menus seem nicer than regular menus in other mp3 players.

Smart Cars

Recently, an Advanced Driver Assistance System (ADAS) hit the news in Europe. I’ve always been interested in advanced navigation systems (and their problems), so I check in on some of the research programs occasionally. After all, individual differences from culture to aging all affect how we use navigation systems.

The original article I mentioned briefly addresses the errors these systems may cause:

Drivers’ uncritical reliance on their sat navs has led to a growing number of mishaps. Last year a woman wrecked her £96,000 Mercedes SL500 trying to drive across a swollen ford through the River Sence in Sheepy Magna, Leicestershire, after her sat nav told her it was a passable route.

…but spent most of the time discussing the errors they catch.

In addition to instructions on when to slow down or change gear for the best fuel economy, motorists will also be warned when they are driving erratically and will even be told at the end of the journey if they have caused undue stress to parts of the car.

Of course, getting to the end of the journey may be more difficult using the current navigation systems. This finding comes from Ziefle, Pappachan, Jakobs and Wallentowitz (2008) who gave an ADAS to older drivers to compensate for age-related perceptual declines. They compared younger and older drivers using either audio or visual aids:

When no assistance was present, driving performance was superior than in both assistance conditions. The visual interface had a lower detrimental effect than the auditory ADAS which had the strongest distracting effect. In contrast to performance outcomes, the auditory interface was rated as more helpful by older drivers compared to the visual interface.

Joe the Plumber as a Persona

There is an interesting and active thread on the IxDA listserv about the potential of using the “Joe the Plumber” character as a Persona in design/usability. First, what is a Persona?

Personas are fictitious characters that are created to represent the different user types within a targeted demographic that might use a site or product. Personas are most often used as part of a user-centered design process for designing software or online applications, in which the goals, desires, and limitations of the user are considered when designing the product. They are also considered a part of interaction design (IxD). Personas are useful in helping to guide decisions about a product, such as features, interactions, and visual design.

Dear Marriott, Snail Mail is Not the Internet

After my stay at the Times Square Marriott a few weeks ago, I received a postcard in the mail asking me to complete a survey. The first 900 people to complete the survey would receive a $10 Amazon gift card. Sounded good to me; I needed some reading for my next plane trip.

However, I soon realized the challenge would not be to be among the first 900. The challenge would be to follow the steps to get to the survey. Below is the text from the printed postcard:

  1. Enter the link exactly as it is printed here: HTTP://SURVEYS.COMMONKNOWLEDGE.COM/I4536SM?RESID=J9B47C2

Well, let’s just stop there. Clicking a link of that length is easy. Typing it into a browser, nearly impossible. Did anyone see the l (lowercase L) in there? Or is it an I (uppercase i)? maybe a 1(one)?

They did provide a way to get help:

If you have any technical issues with the survey link please send an email using the following link: http://surveys.commonknowledge.com/SurveySupport/?sid=4456

And yes, the second link was in lowercase compared to the first link in uppercase, and yes, it was underlined. If only I could click on the postcard. No, there was no phone or alternate method listed.

It’s pretty obvious they just printed postcards with the exact same message they approved for emails. It’s a beautiful case of the message not matching the medium. They could match, if Marriott purchased a simple domain name such as www.MarriottSurvey.com that forwarded to the actual survey, or just included a real email rather than a link to a form. (I’d rather type SurveySupport@Marriott.com).

Reminds me of the video Rich posted earlier.

Another tool to keep up with election news

TechCrunch posted a link to another tool to keep up with the flood of election-related news coming from news services and each of the presidential campaigns.  Dipity Election Center presents news items in a time-line format.

The interface is very cluttered and not exactly intuitive (e.g., unclear what all the little icons below the timeline mean), but an interesting way to present news for news junkies like me.  It presents less at-a-glance information than the website everymomentnow.

How do you keep up with election news?  (if you do at all).

Homeland security warnings

Found a link to this parody site of the homeland security symbols (via AskMetaFilter). This is my favorite symbol:

Michael Jackson is a smooth criminal

The images are real homeland security symbols.  In all seriousness, my friend Chris Mayhorn from North Carolina State University has researched these homeland security symbols and found that not everyone can interpret the real meaning of these warnings:

Mayhorn, C. B., Wogalter, M. S., & Bell, J. L. (2004).  Are we ready?  Misunderstanding homeland security safety symbols.  Ergonomics in Design, 12, 6-14.

Digital TV Transition will be rocky for some

By February 2009, all over-the-air television broadcasts in the United States will be digital.  There are good reasons for the switch such as better use of bandwidth.  However, people who still use rabbit ear antennas for TV reception will need a new digital converter box which is not a simple undertaking.  While this video is obviously tongue-in-cheek, the switch-over will not be easy for many.  As the video illustrates, it is not clear how to get an antenna, and afterward, how to install it.

The VP Debates: An Adventure in Clutter

Perhaps it is because I associate CNN with Atlanta, a city dear to my heart, that I care so much about how badly they choose their on-screen visualizations. Last night I watched before, during and after the debate, which meant I was as informed as could be about their graphics (and even saw the gratuitous use of this Minority Report touch screen.)

This time there was one focus group (32 people), helpfully labeled “Ohio Undecideds” divided into women and men. CNN did apparently fix the scale problem from the first debate, where the reaction line never changed. This time we even saw some ceiling effects:

(Or, I suppose, these Ohio Undecideds were much more polarized than those watching the first debate.)

Now, obviously, I’m a big believer in sampling. However, when they interviewed the 32 people after the debate, it seemed pretty clear that they were mostly registered Democrats or Republicans who said they were undecided so they could be on CNN.* This was allowed almost 1/5th of the total screen during the debate, and was present during the entire debate. I suppose when another large square of the screen is dedicated to distractedly flashing “Vice Presidential Debate”/”Debate Night in America” we can no longer pretend to hold television (even when it the purpose is presenting information) to any rules (e.g., “frequency of use“.)

Continuing my curmudgeonly gripes about on-screen graphics, “points” from six political analysts were displayed on the screen in what I first thought were pie charts, but soon was not sure what to think.


This photo was taken early in the debate. The math became far more difficult by the end

I’ll outline the rules. Analysts could give positive or negative “points” to each candidate when they made a statement. Unlike well-trained Olympic judges, even the analysts on the same ‘side’ were wildly all over the board (some giving a miserly 1-4 points and others apparently madly pressing their button as if they were on Jeopardy.)

My main issue was with load and changability: To get a true number from any analyst I have to subtract the negative points from the positive one. For example, in the picture above, Begala has Sarah Palin at 0, though my first inclination was to add those two numbers. If I want to get an idea across analysts, I have to do that for each one, hold the final number in memory and move to the next (or switch from adding to subtracting as I move down and across the screen.) Then I have to remember the final number for a candidate and process the next candidate.

Worse, the analysts could change their already given points, making a mess of the idea of +/-. Just because one candidate had 3 positive points did not mean they wouldn’t have 2 or 1 the next time you checked. This undermines the idea that they are getting additive positive scores for good points they make and negative scores for incorrect, unpopular, or lame points they make. If that were the case, each number should only increase throughout the debate.

Perhaps it was the large amounts of pizza and chocolate consumed at the debate party, but once all analysts reached double digits, I gave up and tried to ignore the flashing numbers.

As another note, many of these analysts had party biases, however their leanings were not noted on the screen. It would have been helpful to have them separated on each side by that pre-debate bias.

Last, my favorite moment came after the debate when someone at CNN printed out the analysts pie-chart-ish results and put them on screen to talk about them while the actual graphics were still up on the sides of the screen!

*I mean no offense if one of the 32 reads this and doesn’t fit that claim. There were a few true Independents interviewed.