Voting should not be a black box affair

The cliche goes, "When all you have is a hammer, everything looks like a nail". As computer professionals, that means that every problem tends to give us an itch to develop a computer-based solution. This is unfortunate, part of an educational process that emphasizes programming skills, technology, and narrow theory over problem solving. When we're confronted with a problem or process, we first need to concentrate on really understanding the existing situation. What is the current goal? What is the problem? Who are the stakeholders? What aspects of the system are important to the different stakeholders? Only then can we take the first step in solving the problem: identifying the appropriate range of solutions, which may not involve computer technology.

A case in point is our elections system. The problems with electronic voting systems are so rampant that I'll not even bother to link to articles on the web. What are their advantages? According to Wikipedia (not that that's the final source on this matter), the primary advantages of electronic voting systems are saved paper, easier accessibility for people with disabilities, and the possibility of faster counting. Wikipedia also mentions the possibility of cryptographic verification, but so few people understand what that is -- likely, many computer professionals don't -- that I wouldn't consider it a practical advantage. After all, how can you expect someone to consider their vote "verified" if they don't understand the verification process? Verification then becomes another word for "trust us; we're experts" or "the magic box says that everything is fine". There's only one problem with any of the above advantages: they have nothing to do with the most important problems of paper ballots.

A recent talk at my school by Bev Harris, founder of Black Box Voting, I think very clearly pointed out the fallacy of a technological fix to the problems of paper ballots. Ms. Harris is not a technologist, and it behooves us to listen carefully to what people like her have to say. She spent much of her talk discussing one horror story after another related to electronic voting systems: conflicts of interest (putting it in the best light possible) with candidates owning interest in companies that manufacture the machines that count their ballots, lax software development security allowing anyone to access -- or even modify -- voting machine software via the Internet, software with back doors, hardware that allows anyone with a little knowledge to pre-load bogus voting information, etc. Other than ownership of electronic voting equipment manufacturers by governments considered unfriendly, these issues curiously don't seem to concern incumbent politicians that much.

And then she made what was to me an eye-opening statement (which I'll paraphrase because I didn't write it down at the time). The problem with electronic voting systems is that they, on the one hand, lack transparency, and thus make it more difficult to detect fraud, and on the other hand, reduce the number of people needed to commit a fraud. The lack of transparency arises in two ways: only a few people observe even part of the process (much of it is automated, and thus completely unobserved) and only a few people have the expertise to understand the different ways the count could go wrong. Electronic voting systems also make fraud easier, by reducing the number of people who need to be "in the know".

The most important attribute of any voting system is that its users -- the general population -- have confidence in its integrity. All other considerations are secondary. Even the presence of fraud is secondary, because confidence can only flow from ease of true verification, which by definition makes it likely that fraud will be detected. I would assert that electronic voting systems can never engender as much public confidence as paper ballots. Any proposed solution to voting problems must take this into account. That's why I'm especially impressed with Black Box Voting's Citizen's Tool Kit, which empowers "ordinary people" to oversee how their votes are counted, to demand that all methods see the light of day, and to generally make themselves a pain in the rear of anyone in government who would rather that the unwashed masses leave the process to the "professionals".

Some problems do not have technological solutions.

Topics: politics, voting, elections, computer security.

Vanity press

Those WMSCI ("World Multi-Conference on Systemics, Cybernetics and Informatics") people seem to be at it early this year, with their solicitations for submissions to their curious conference. Click the title above to learn more about how to save money, your reputation, and put your work in a more useful venue. If you want to feed your vanity, start a blog like everyone else!

Which American President Are You Most Similar To?

Via Living the Scientific Life, the President I'm most like:

You Are Most Like John F. Kennedy

You live a fairy tale life that most people envy. And while you may have a few dark secrets, few people know them.

What Modern US President Are You Most Like?

Studicious

No, it has nothing to do with Congressional Pages. Via Brad DeLong's Semi-Daily Journal comes a early version of a web site (linked from the title above) that allows user to enter and share class notes. It's an interesting idea, and it makes me wonder about the possible uses of collective note-taking. A substitute for taking -- and rewriting, with refinements, while studying -- notes it is not.

Neat mapping tool

I ran across Gmaps Pedometer as a way to measure the length of my bicycle commute from home to work (via bikejournal.com. It combines the Google Maps interface with the ability to double-click on the map to set markers for your route. It also includes the ability to show an elevation profile of your route. So now I know that I have a 20% hill climb near the end of my ride back home (200ft. elevation gain in 1000ft of cycling).

State testers fail math

wasl.png
Sample WASL bar chart, originally uploaded by stiber.

My older daughter just received her results from the WASL (Washington Assessment of Student Learning), which is this state's test that is actually used to assess schools. I have ethical issues with requiring students to take tests that don't benefit them. In the case of the WASL, all that students receive is their scores; they get no feedback that they could use to identify weaknesses, nor does the score impact their education beyond the requirement that they spend time studying for the test.

But that's not the topic of this post. Instead, I'd like you to look at the sample bar chart, taken from this annotated example. It's not too dissimilar from what we received for our daughter, except for details such as her scores and comparisons with her school, school district, and state averages.

Look at the vertical scale for this bar chart. The first thing we notice is that the Y axis doesn't start at zero, yet there is no indication in the bars that their bottoms are cut off. Maybe we could dismiss this; perhaps the purpose of this graph is to show differences (more obvious in what we got, which has multiple bars). But look at the numbers on the Y axis: each of the four categories takes up the same space on the axis, even though their ranges vary widely, from a minimum of 25 to a maximum of 275!

In other words, the vertical scale provides no useful information, and in fact is misleading in implying that all four categories have similar ranges. Not only that, but by sharing ranges between reading and math, the fact that the ranges for the same categories differ is obscured.

So, the heights of the bars are meaningless. Bars with tops in different categories cannot be compared because of the nonlinear scaling. Bars in different tests cannot be compared. About the only information conveyed by this graph is the category that the student in question's score fell within, and possibly the relative position within that category. This begs the question: if the folks who design the WASL report can produce such a bad graph, how confident can one be in the rest of the testing and scoring process?

Topics: education, testing.

Travel photo of the week

Guilin cave
Originally uploaded by stiber.

In the US, the general approach to providing public access to caves (these days) is to set things up so that the impact of visitors is minimized. The focus of the tour is on the science (geology, sometimes biology) of the place.

In China, caves are used to tell stories, like you might do when looking up at the clouds. Tours focus on historical and mythological figures and events, as if the rock formations were dioramas. Alternatively, areas are lit up like a light show. The cave experience becomes a cultural one.

Bicycle commuting

There comes a time in everyone's life when you must decide: exercise or start saving for coronary bypass surgery. I've been running more, but sometimes things are just too busy for me to fit a run in. That, plus my previously-voiced frustration over the lack of progress in car fuel efficiency by automakers (despite existing modifications that can yield 100 miles/gallon), led me to think seriously about bicycling commuting.

I've ridden to work before, and do so when the weather is nice: one way. After work, my wife would then pick me and the bike up. The main reason for this is that we live at the top of one of the (if not the) tallest hills in the area. I decided that I would see if I could ride up, so that I could commute both ways by bike, and in fact it seemed feasible when I tried. I'm in OK shape, but from running, not biking, and only had to walk two half blocks (the steepest parts of the route -- steep enough that you might wonder if your car can make it). This leads me to believe that I'll be able to do the climb without stopping one day, and it's always nice to have a goal to shoot for.

Next comes the strategy to avoid procrastination and backsliding ("I'm a bit tired today; I'll bike in tomorrow," where tomorrows eventually become months). Part of this involves buying equipment to set my bike up properly. I've got a 7 or 8 year old Raleigh R300 touring road bike; I outfitted it with fenders, added a flashing tail light, and am in the process of getting a good, bright headlight (I already had a rear rack and panniers). I also bought tools (I used to do all my bike maintenance, but got rid of all my tools after grad school). Now if I don't commute, I'll feel like I wasted all that money. I also estimate that we'll be saving about $20/week on gas (the commute is a bit over 5 miles, one way, we only own one car -- an all-wheel-drive Dodge Grand Caravan -- so my wife often drops me off at work in the morning and makes another trip in the evening to pick me up; it adds up). I get about an hour or so of exercise per day, broken into two parts each of which increases my metabolism for an extended period of time afterwards.

I found an enormous number of bicycle commuting resources on the web. I'll just point out one: Paul Dorn's Bike Commuting Tips.

P.S. If you're biking home from work and it's uphill all the way, eat a snack beforehand.

Topics: bicycling, transportation, environment, exercise.

On the value of computers for primary education

Daniel Lemire commented on a previous post of mine, and I thought I had an elaboration that merited a post, rather than burial in a comments thread. I had written that computers could be useful as tools in primary school, but not as objects of study themselves. He suggested a parallelism with pencils, etc., and yes, those other things are tools, too. But, while those are necessary, I don't believe that computers are necessary tools. Potentially useful, yes, but kids can learn the same things without computers. Here's a quote by Alan Kay from "Revealing the Elephant: The Uses and Misuses of Computers in Education":

Suppose it were music the nation is concerned about. Our parents are worried that their children won't succeed in life unless they are musicians. Our musical test scores are the lowest in the world. After much hue and cry, Congress comes up with a technological solution: 'By the year 2000 we will put a piano in every classroom! But there are no funds to hire musicians, so we will retrain the existing teachers for two weeks every summer. That should solve the problem!' But we know that nothing much will happen here, because as any musician will tell you, the music is not in the piano!

Garden photo of the week

Tomatoes
Originally uploaded by stiber.

We've had a pretty successful garden this year. Well, the fava beans didn't do that well this winter (probably due to competition from weeds that were a product of my sabbatical), but the tomatoes have done fantastically. This is a non-trivial matter, as anyone familiar with vegetable gardening in the Pacific Northwest will confirm. The big, red ones are Siletz; many are of respectable beefsteak size (I'm very particular about such things, having grown up eating New Jersey beefsteaks purchased at produce stands in Philadelphia). Soon, it will be time to put in the fava beans again, and maybe we'll try some collard greens.

A computer for young children

I recently underwent an adventure searching for a suitable laptop for my daughter that took me back to my past. Along the way, I was reminded that the latest technology is not always the greatest. So, what do primary school children need in a laptop computer?

The primary use of a laptop is as a way to write and produce "nice" output. It's remarkably motivating for a child to know that she can make her writing look just like a "real" article or book. In my daughter's class, they use old versions of the AlphaSmart laptop. It's a keyboard with a one-line LCD display and rudimentary software that lets children "edit" up to eight files, each recalled by a function key. It interfaces with computers via an ADB (Apple Desktop Bus) port, and via an ADB to USB interface cable for new computers. The AlphaSmarts allow students to each have a keyboard to type in text without the school needing to buy so many desktop machines. I think this is both reasonable and a prudent allocation of school resources. Computers can be useful tools for some specific activities at the primary school level, but they should not be the focus of class time.

So, clearly the AlphaSmart is fairly basic. The newer models are fancier, but as you can see from the web site, they are still limited, they run a proprietary OS that requires the purchase of additional software from the manufacturer, and they cost $400! Given the recent project to develop a $100 laptop for education, this seems excessive.

What does a young child need in a laptop? She needs durability (so no hard drive or thin lid with large display), reasonable battery life (so no hard drive), simplicity (so the OS needs to be hidden), a keyboard (no palmtops), and basic connectivity. It occurred to me that all of these things were features of my old Newton MessagePad 2100, except for the keyboard. But there was another machine using the Newton OS: the eMate. And the eMate fits this task perfectly, including the simplicity and power of an OS that is still unmatched today. This includes instant on, no need to explicitly save documents, the ability to install additional software, non-volatile memory (so there's no danger of you ever losing a file, unless you delete it), stylus and keyboard, etc., etc. Others have commented on using the eMate as a child's computer. You can get an eMate on eBay for less than $30, including shipping costs, which is a nice feature, too. The only question was: could I get it to work with today's computers?

Power
Some of the eMates on sale on eBay will have power supplies and some won't. The one I bought didn't; a Radio Shack 3-7.5V, 2A digital camera power supply (catalog number 273-1696) works fine. Though nominally the eMate requires 7.5V (the original power supply was rated 7.5V, 1.2A), I found that the eMate complained about 7.5V but was happy with the power supply set to 7V.

Batteries are another matter. Odds are, the eMate you buy will have an original NiMH battery pack that doesn't hold much of a charge. You can either replace the cells in the batter pack or replace the battery pack with a holder for 4 AA batteries. As of this writing, I'm planning to do the latter.

USB Interface

The eMate has a built-in serial connector. I had a Keyspan USA-19Qi serial to USB adapter from an old Palm that I'm no longer using. However, the eMate's serial port uses a Apple mini DIN-8 connector, while the Keyspan uses a PC style DB9 connector. I had some Apple style and PC style cables and was able to make a DIN-8 to DB9 cable, given some information about the pinouts of the two ports. No doubt, I could have bought something somewhere, but I had the hardware and expertise to do it.

Mac OS X Software

Here you're in luck, as there is a (somewhat) active developer community out there for the Newton OS (which says something about the hardware's longevity). The software I chose is NewtSync, which aspires to be an iSync-like program for the Newton. It requires software on the eMate side, too, which can be installed using the NewTen package installer. With the TextSync plugin, my daughter can write papers using the eMate Notes program and then sync the notes to a Mac running OS X. Works like a charm. With other plugins, a Newton can be a perfectly usable PDA.

Topics: computers, children, education, newton.

[+/-] show/hide the rest of this post

Getting things done

I reached my limit this summer: too many things to do, too many emails in my inbox, too many meetings (I'm now Vice Chair of our faculty organization; basically the campus senate), and too many things not getting done. And this is summer; it will only get worse as the school year starts. So, thanks to Merlin Mann's 43 Folders web site, I got turned on to Getting Things Done: The Art of Stress-Free Productivity, by David Allen. The author's methods are great for me, focusing on workflow practices that can support better productivity. More than just being productivity focused, Allen emphasizes practices that remove stress from one's life, by removing the need to remember all the little things you need to do (it's all in the workflow system) and allowing you to determine what you can't do (thus removing them from your system and out of mind). His approach builds in the idea of context: labeling each task by the location or work mode in which it can be done, so you're never looking at tasks that aren't possible to do right now. I'm still in the process of grinding through my many piles and files for tasks to enter into the system, but I highly recommend this book.

One thing that has helped me implement this approach has been some very useful software. You can look through Merlin Mann's web site, or the lifehacker web site, for the tools that fit you best. I've found three tools that work for me:

1. I like simplicity, portability, and universality. That's one of the reasons I like LaTeX for writing: all of my files are plain text and so I never have to worry about having stuff trapped into obsolete or incompatible formats. That's why I love the todo.txt software, which manages plain-text task lists. There's an increasing amount of community-contributed software that provides various capabilities, including my own software for managing lists of project-related tasks.
2. GeekTool lets me put the output of various todo.txt programs on my desktop; it also allows me to quickly switch the display for different work contexts.
3. A great app that implements a "tickler" system in Apple's Mail program is GTDMail. It is an Applescript that creates a set of tickler folders: one for each day over the next month and one for each month of the year. You set up iCal to run the script each morning, and it moves the items from today's folder(s) (the current date and the current month) to a "@INBOX" folder. I modified the target folder to "@ACTION" and added "@READ-REVIEW" and "@WAITING FOR" folders. So now I can process all of my emails within seconds according to Allen's workflow.

I feel better already.

Topics: personal productivity, book review, organization, software.

Astrologers rescue Pluto

I can't tell you how happy I was to see the front page article linked from the title above in today's Seattle Times. However, I was able to express myself to the newspaper's editors:

To the Editor:

Thank you very much for your much-needed front page injection of superstition and ignorance into the scientific debate over the modern definition of the word "planet" ("Boot Pluto? Astrology's dismay off the charts", Saturday, August 26, 2006). The sidebar listing astrology "facts" was an especially nice touch. What's next? Reaction from bloodletters on the latest cancer treatments? You do your profession proud!

Sincerely,

Michael Stiber

US Department of Education vanishes evolution

The title above links to a New York Times article describing how a US Dept. of Education grant program in which evolutionary biology is conspicuously missing. You can look at the document in question here; evolutionary biology should be CIP code 26.1303 on page 7. Instead, there is a blank line. DOE spokespeople are saying that it was a clerical error. The interesting thing is that there are some places in the document (such as where evolutionary biology should be) where a number is skipped and there is a blank line, and there are some places where a number is skipped and there is no blank line. It appears that, in the latter case, the skipped CIP codes don't exist, while the blank lines in the former case indicate that a CIP code was purposefully deleted (though there are some blank lines where there is no CIP code). There are only two blank lines that correspond to deleted codes: 26.0908 (Exercise Physiology) and 26.1303 (Evolutionary Biology). (You can look up CIP codes here.) If we assume that these edits both were done in the same fashion, then this would seem to support the idea that evolutionary biology was purposefully, not accidentally, deleted. Topics: evolution,creationism,politics.

Travel photo of the week

Farm and hills
Originally uploaded by stiber.

For those unfamiliar with the area, this is what the South Dakota badlands might look like if the climate there was wet and warm.