Ethnography Considered Harmful

Last in a series of "Considered Harmful" papers is Ethnography Considered Harmful by Andy Crabtree, Tom Rodden, Peter Tolmie from the University of Nottingham and Graham Button from Sheffield Hallam University. This paper argues that Computer Scientists have taken the ethnography to new and dangerous direction that skew the results of their analysis. I think that any analysis technique can be skewed, but as computers become more and more a part of everyday life, they become the same as any tools that any ethnographer has ever written about, and can then be used in many new ways of analysis. I do however agree with the paper in stating that design ethnographies should not jump to conclusions on rhetoric or deep analysis of actions to create a new design. Ethnographies should serve more of a purpose to find flaws in how people use computers today and how to fix them.

Human Centered Design Considered Harmful

Another paper in the "Considered Harmful" series is Human Centered Design Considered Harmful by Don Norman. In this paper Norman argues against Human Centered design and instead favors Activity Centered Design. Human centered design states that all computer activities should mold their form around the human, and designers should try to understand the human behind the interface. This, however, doesn't take the action into context and this is exactly what Norman points out. Norman points out that too much focus on the user can lead to a complicated software design and makes the software more difficult to update and create.
I agree with most of what Norman says in this paper, but believe that perhaps a balance between Human and Activity centered design is the ideal. The focus should be based on what part of the software is being designed and what is necessary at the time. This of course is a very path to walk down and is not easy to find.

Usability Design Considered Harmful

This paper is titled Usability Evaluation Considered Harmful (Some of the Time) and is written by Saul Greenburg of the University of Calgary and Bill Buxton from Microsoft Research. Their paper reflects on the recent upsurge of usability evaluations in CHI papers at every level of development. They repeatedly state that they do appreciate the benefits of usability evaluations at the end of a research project, but early in the creation process they adamantly oppose them. The CHI field is full of new, strange, and novel ways to interact with computers, and this paper suggests that evaluating these young technologies against long-developed, traditional methods of interaction may prove that the new technique is inferior to the traditional one, even though the new technique is just not fully developed.
I agree with the statements of the paper and have read several research papers were usability evaluations seemed to be tagged on as an afternote to appeal to the conference committees. Usability evaluations should be considered as a possibility depending on the nature of the research project and not a necessity.

Fitts's Law

This blog post is about a paper written in 1954 by Paul M. Fitts. This paper, titled The Information Capacity of the Human Motor System in Controlling the Amplitude of Movement, contains the description of what is now called Fitts's law. In the general case it shows that how much time it takes to perform a task is based on the distance something must travel to perform an action, and how accurate the action must be. This law has proven to be near infallible over the years and is an excellent way to analyze efficiency of a task.
While the paper was math-dense and the descriptions were very general, I think it is these things that keep Fitts's findings alive even now. One thing I am very curious about is an analysis of various keyboard configurations using Fitts's law to see how they compare.