http://aaronkirkes-chi.blogspot.com/2011/01/reading-1-only-one-fitts-law-formula.html
http://wkhciblog.blogspot.com/2011/01/paper-reading-1-only-one-fitts-law.html
Reference:
Only one Fitts' law formula please!
By: Heiko Drewes
Venue: CHI 2010: I Need Your Input, April 10–15, 2010, Atlanta, GA, USA
Summary:
Fitts' law correlates the amount of time it takes a user to click on a target with the distance the cursor is from the target. However, the HCI community has adapted or modified the original formula to best fit their data without any reasoning for doing so. Furthermore, Drewes says that the HCI community allows this to happen without ever truly questioning the reasoning behind using different versions of the formula. He believes the multiple formulas are a contradiction and do not represent proper scientific studies. Some of the versions of the formula try to eliminate the possibility of negative values, but Drewes explains that the only way the formula can return a negative is if the cursor was already on the target, meaning in that position in the past. He states that because of this the original formula was correct and there was no reason to deviate from the original formula. Finally, he concludes that the HCI community should either treat Fitts' law as an art, which would allow contradictions, or a science, which would require one formula to be chosen.
Discussion:
This was a pretty confusing article. Drewes presents a problem without ever truly answering it. It seemed to me that he mostly wrote this to criticize the works of other people. The best answer he proposes to the problem he was aiming to answer is that the HCI community should decide whether to treat Fitts' law as an art or a science, but not both. Basically he's passing off the entire issue to everyone else, when he's the one complaining about it in the first place. I just didn't see the purpose in this article other than to bash other works and thus pointing a finger in their respective faces.
No comments:
Post a Comment