In another post I describe “demandance”, a theory that, in small part, describes how to design (or not) devices that are usable by young children without being distracting. In the demandance paper I give the example of a great Apple game called AlphaBaby. The specifically interesting thing about AlphaBaby (at least for the present post) is that anything that a child would possibly type, does something interesting! In order to get out of the app, you have to either kill it using the operating system interface, or type the letters Q U I T in that order. And, by the way, the Q U and I in that sequence do the normal AlphaBaby thing, so unless you know to type those letters in that order, you can’t get out of the app. What all this means is that the child pretty much can’t get out of the app, which means that you can leave him or her there, and walk away, and not worry about son or daughter either getting frustrated by dropping out of it, or, even worse, dropping out of it and then screwing up something in your “real” environment.
To reiterate the principle that I want to draw from AlphaBaby: Make it as easy as possible to do the right thing, and as hard as possible to accidentally get out of the state in which the right thing is talking place.
Okay, so on to image search. Image search (for example from any of the major search engines) is a terrific tool for teaching typing and spelling; you type in something you’re interested in, and get back a bunch of cool pictures. What could be more motivating? (Well, perhaps actually getting the thing itself … someday! 🙂 ) Unfortunately, these tools run in browsers, and so are burdened by the clunky browser interactivity model…not clunky for you, perhaps, but definitely clunky for a child. The worst part is that you have to actually know how to use the mouse (or touch screen) and remember to select the search text box. What you really want is more like AlphaBaby, so that anytime you type, it searches; no fumbling for the mouse or trackpad and trying to click into the search box. (You’d think that pads would be better at this because you can just click instead of having to use a mouse, but in fact, they’re worse because of the pop-up keyboard, which is basically a whole separate modal interaction.)
So I created a simple wrapper on Google Image search, called “ColorPix“.
When ColorPix comes up, it looks like the screen on the left, with “RED TRAIN” automatically entered at the top. You can page through the results using the arrow keys. Now, the interesting part, and the only novelty here — but I claim that it’s a big one! — is that by simply typing, it starts re-writing what you type into the text field at the top. So to get from the left screen to the right screen required no clicking or tabbing, I just typed the letters F U N N Y <space> F R O G (and pressed ENTER — the ENTER could be made optional, but if you implement iterative continuous updating you’re going to find yourself in trouble when you son or daughter intends to type: D I C K <space> T R A C Y, and only gets through the first four letters! 🙂 )
Anyway, so ColorPix is really simple — definitely not perfect. But in my opinion it gets image search, which is very fun and useful for teaching typing and spelling, a little bit closer to the desiderata of good interactivity design for children.