Vote for green…I mean red – Part 2 / or Freedom 2003
Lab Report February 11, 2009
On Wednesday we were coming in and Matt took us by some surprise. He announced that we start with the presentations of the biased voting machines and have our talk about the readings afterwards.
We had 30 minutes (which was extended to 90 minutes – Puh!!)
We were facing a problem considering the fact that we had nothing besides an idea, some blown out three-coloured LEDs from last week, and an empty Arduino board.
So we had to speed up and do some compromises. Well if the handling of the three-coloured LEDs is so complicated for us we would have to live with two LED (one red and one green). These two LED are simply connected to power with an resistor in between to prevent them from blowing up. The rest of the circuit is to read the pressing of the buttons on two digital INPUT pins. This is a very simple circuit that we did at the beginning of the course. Considering the time pressure it was the only thing we were able to do in 90 minutes.
The major problem we were facing is the fact that Jamon (our colour-blind test person) could still distinguish the red and green LEDs. The solution was to have a yellow background as seen in the following picture.
It turns out that colour-blindness is depending on the context (and might vary from person to person). Jamon could tell the LEDs apart with the desk as a background. Once placed in the voting machine (cardboard box wrapped with yellow tape) the yellow background made the LEDs indistinguishable for Jamon.
Next to each LED we placed a push button and count the number of times it is pushed. We had a debug output on the serial console telling the number of votes for a red or green team. In a biased environment we would ask to press the button besides the red LED to vote for A and besides the green LED to vote for B.
There are several biases built into our voting machine. The one against colour blind is very obvious. Another bias would be to give no feedback at all. Even if the voter has normal sight s/he will not be able to tell if the vote was counted correctly.
The cultural bias built in is the fact that we could ask a yes/no question and associate red with yes and green with no. This is no big deal and still very uncommon in our society. Furthermore, the bigger button with the red light might attract more attention.
Our voting machine is deeply biased and from the outside a black box that does not allow the voter to understand the processing of the votes. This is a desired result of the design process. However, many existing voting machines represent a black box since they are produced by a company. The insides are a trade secret of the company manufacturing the machines. This is a common problem with proprietary products and designs.\
The question is if we want to understand how our technologies work or if we just take these technologies for granted. This might be one of the goals of this course – learning to critically look behind the scenes without becoming to technological.