Double Click Duration

9 May 2010 - 4:52pm
6 years ago
2 replies
1570 reads

Hi folks,

I'm trying to dig up some guidelines for implementing double-click.  I can and have searched teh internets and consulted, e.g., The Essential Guide to UI Design. :)  I also know most of the principles and can derive from what the OSes and such do as well as the next person.

I say all that in hopes of focusing replies--I'm specifically looking for references based on testing and/or real-world practice.

I'm interested, in particular, in good defaults for interpreting double click in platforms that do not handle that for you via, e.g., the OS settings.  TiA.



9 May 2010 - 7:05pm
Santiago Bustelo

Hi Ambrose,

I have not read "The Essential Guide to UI Design". From what I peeked, it looks very valuable, and I have placed it next on my read list. Here are my two pence - hope it actually brings some new information to your quest:

A double-click is a problem of both time & space. Let's go with time first:

After receiving the first click, the system may need a time window to determine if that was the only click, or if it was the first click in a double-click gesture. If that is the case, it is good to know how much time a double click gesture will need according to KLM-GOMS model:

Mouse down: 0.1 seconds

Releasing: 0.1 seconds

Mouse down: 0.1 second
Releasing: 0.1 seconds

The time window between the first and the second click is 0.2 seconds (first releasing + second mouse down). That is just below the perception threshold of 250 milliseconds. So far with the good news.

These measures have to be adjusted using a time multiplier according to the age of the operator. The highest multiplier is 2.2 for operators above 65 years old, giving 0.22 seconds for each operation, and 0.44 for the time window, a measurement that will be uncomfortable for younger or more skilled operators.

A solution would be to start using the more generous value, and to dynamically adjust the time window to the operator's speed.

On the space department:

The implementation should take into account that a small mouse motion would be inevitable between the two clicks. The user will click the same object twice, but not necessarily the same pixel.

Fitt’s law may be of some use to measure up how much the cursor could travel in a *voluntary* motion on the time required for the first release, allowing to tell apart involuntary motion. I have the gut feeling that following this approach would be an overkill, as just thinking about it made my head hurt ;-)


Santiago Bustelo, Icograma
Buenos Aires, Argentina

// IxDA Buenos Aires:


17 May 2010 - 1:46pm

Thanks, Santiago! 

I just got back from some traveling, hence my belated reply.  I guess it would be safe to say that the target they're clicking on would be the space/area to count as multiple clicks on the same thing.  Meaning, if they click twice in the timespan on the same object (button, paragraph, whatever), it's safe (usually) to count as a double click.


Syndicate content Get the feed