When Usability is used the wrong way
First off, let me commend Tolerance.org for keeping this subject in the foreground. They make a lot of good points, about how even if we consciously believe we have little or no bias against other races, genders, or people's age that they still lurk below the surface.
The site offers a number of 'tests' to demonstrate how much of a bias you may have unconsciously influencing all of your interactions with other people. This is where I have a bit of difficulty with their message.
Each of these tests involve quickly viewing a picture or word and pressing a key to say the image is good or bad, black or white, young or old, etc. The two test that I have taken (race and age) present a series of pictures and words and you press 'e' or 'i' to identify what they show. For example: Images of faces are shown and you press 'e' if the face is old and 'i' if it is young. Then they present words: press 'e' for negative words (pain, anger, etc.) and 'i' for positive words (happy, joy, etc.). Then they combine them through two series: Press 'e' for Old faces or negative words, press 'i' for young faces or positive words. For the last series they switch the keys: Press 'e' for young and positive words, press 'i' for old faces and negative words. You are asked to go very fast and not think too much about the images.
However, by going through so many series where one association is made (i.e. old and negative) with a particular key on the keyboard, and then switching the keys, it is no wonder people demonstrate much stronger biases than they think they have. Interacting with a computer in this way trains the response to happen from the spinal column and not the brain. For example, if you know how to touch type, how much faster do you type when you do not think about where the keys are as opposed to thinking about which key needs to be pressed to spell a word. I cannot say how many times during the test I knew in my head that I wanted to press either 'e' or 'i', but my spinal column reacted from the training before my head could stop my finger from pressing the wrong key. I am sure they think that this reaction proves their point. Let's see how they do if we switch the brake and gas pedal on their cars and then have a 3 year old run in front of them. The results would not be pretty. They key here is the repetition in the first four passes of the test; you have been trained in that short amount of time to draw an association between a particular image and a key on the keyboard.
Again, I applaud tolerance.org for their stance, but I really do not think their 'tests' demonstrate a bias, but the trouble with Human-Computer Interaction.
8:02:35 AM 
|
|
