Fitts' law and minimum target size

You can read everywhere that bigger targets are easier to hit. But say I have a given input modality and want to know the minimum button size for the interface. How would I go about that?

Fitts' law says

``````T = a + b log2(1 + D/W)
``````

So I guess I could define a maximum T knowing a and b, but what about D ? How can I express some kind of minimum for the interaction time?

• Don't use Fitts law to settle your minimum target size guidelines. Use real users instead. Aug 9, 2012 at 14:46
• The constants in the expression above are "to be determined empericaly". So, use the equation as a way to understand how distance and size influence the time it takes to aquire a target, but don't try to use it as a way to calculate times or sizes beforehand. Aug 10, 2012 at 11:36
• @JørnE.Angeltveit I think the idea of the law is that it will work for real users otherwise it would be pointless Aug 13, 2012 at 8:54
• @André a and b are to be determined empirically since they depend on your input modality but T/D/W are really the things you are trying to determine as i understand it. Aug 13, 2012 at 8:56
• @joekr. No! The idea of the law is not that it should work with real users! And no, the work Fitts did wasn't pointless - of course it wasn't! Fitts did some basic and elementary research on human movement. One of the outcomes from that research was the popular formula, but that's not the important part of his work. The formula is a model. And just like any other model, it's a representation of the real world - it is not the real world! Have you ever seen any of the scatter plots that "makes" this law? It's mess. Real users are so unlike the formula. Aug 13, 2012 at 9:49