As many engineers do, I sometimes help out computer novices of varying degree with questions or problems. The last couple of times I did this I noticed a point of confusion that is rarely discussed. I saw the people I was helping repeatedly click the wrong number of times on a GUI element. Sometimes the desired effect was still achieved, and the person didn't seem to know that they were doing anything wrong. Other times it did cause a problem, but it wasn't necessarily clear to the person that the number of clicks was the cause. Sometimes the person did ask me afterward about how many times they need to click, and I would answer. Whenever a question like that comes up I try to provide an overarching answer in the spirit of teaching someone to fish rather than giving them a fish. But after thinking for a few seconds I couldn't quickly come up with a consistent rule about when to click and when to double-click, to say nothing of more obscure maneuvers. It also suddenly struck me how little attention this question gets as compared to, say, the debate over the number of buttons that should be on a mouse.
As an exercise, then, I'll now try to mentally go through all the types of clicking I do on a daily basis, and write out which types require a double-click, since one would think that should be the less common action.
- In Finder or Windows Explorer, or on the Desktop, double-click an icon to open the file, folder or application.
- In a browser's location bar, and in Microsoft Word, double-click to select a word, triple-click to select the whole line.
- In email applications, click a message once to display it in a panel, double-click it to open it in its own separate window.
Those are the first few that come to mind--clearly it gets more complicated as you get further into the land of third-party software and its interfaces. The explanations above also rely on some concepts that many people don't understand. I'm realizing that there are probably a lot of people out there who use computers daily for email, web browsing and other tasks, and who don't know what Windows Explorer is, or how to find a file on the hard drive. A lot of the clicking errors I've seen were double-clicking on links in web pages or on "OK" buttons when one click would suffice. Sometimes this causes problems when the user ends up clicking once on a GUI element that was displayed or revealed after the first click. This practice probably develops in the user after frustrating experiences with clicking and seeing nothing happen.
From a designer's perspective, it makes sense to use double-clicking only when we need two different response behaviors that are both based on clicking on something. When we do so, it's good to think about how the user will know, other than by trial and error, which number of clicks will produce which response. We can usually learn keyboard shortcuts by exploring an application's menus, where the key combination is displayed next to the description. But there's no analogous visual indication of what mouse clicks will do1, and instead we (as users now) must employ the unconscious knowledge that some, but not all of us have about the 'rules' of computing and graphical interfaces.
[1]: In this way, right-click menus are better, and allow for any number of actions to be associated with clicking. They do require more effort than a single or double click, but the double-click action could be listed in the context menu along with the right-click actions.
