http://stuartjchi.blogspot.com/2011/01/conference-paper-17-only-one-fitts-law.html
http://tamucsce436-2011a.blogspot.com/2011/01/reading-1-only-one-fitts-law-formula.html
Reference Information
Title: Only One Fitts' Law Formula - Please!
Author: Heiko Drewes
Publisher: CHI EA '10, April 10-15, 2010
Summary
Fitts' Law essentially models a relation for the average time it takes to point at a target that is a given size and distance away. The model shows that the farther away or smaller a button is, the longer time it takes to click it. This form of pointing is very important to the field of HCI because it is a very frequent action that users perform when interacting with computers.
The author states that there are a "high number of publications" in regards to FItts' Law, and it seems that "the big amount of research papers contribute more to confusion than to a clarification of the topic". The discrepancies can be traced back to misunderstanding and conveniently tweaking certain variables as to show correlation when one may not exist. Since Fitts' Law is one of the very few quantifiable pieces of data in the field of HCI, it would be good for everyone to use the same formula.
The author goes on to say that while Fitts' Law is important in that it can model the system, in reality, button placement is kind of common sense. Buttons should not be too small or too big. "The task of HCI is to find a good balance and this can already be achieved with common sense."
Discussion
I found this article to be very telling of the state of HCI. It seems that there is no true standard when it comes to Fitts' Law if there are so many variations and justifications for such discrepancies. I like how the author stated that if they [the community] were to truly acknowledge what is right and wrong, it would render huge amounts of HCI research "obsolete." Instead, Drewes proposes that HCI be treated as an art. The community would "live with contradictions and more than one truth." Art cannot be quantified, and HCI could go that route. However, if it wants to be considered a science, it must be held to certain scientific standards.
I felt like there were a lot of technical terms in this paper that were never defined, lending to my confusion when it came to formulas and such. I never figured out what an "ID" was.
Image: http://www.clker.com/clipart-1795.html
I felt the same when reading this article about the terms not being defined. From reading this article it sounds like the HCI community is already treated as an "art" and that's why using all four different variations of the formula is acceptable, but if it wants to considered a science they should ultimately go by scientific standards.
ReplyDeleteI never figured out what ID was defined as either unfortunately. I think that there's nothing wrong with HCI being treated like an art, but if people are going to continue to use Fitts' law, it should be defined.
ReplyDeleteJust like these two guys, I feel that despite the technical issues marring the paper, this argument is good enough that it merited publication.
ReplyDeleteAs far as the argument between art vs. science, I feel that despite the fact that HCI can be a matter of opinion at times, since it has a strong link to engineering, it should follow scientific principles, and thus they need to stick with a single formula.