I've been teaching middle school kids programming for four years. I don't think the language is that important. What's important is to find good problems that they can work on. And by good I mean: (1) Students find them interesting, (2) early problems are easy so everyone can get a small victory under their belt, (3) the problems scale up so everyone has something that challenges them.
The way I teach is I start everyone out on Scratch ( scratch.mit.edu ). Scratch uses drag-and-drop programming which lets you skip all the worries about syntax and jump immediately to logic, problem solving, etc. I.e. the interesting stuff. Most of the students will continue to find Scratch interesting and challenging for a long time (6-months to a year). Your top kids will get bored and then I'd move onto something like Processing ( http://processing.org/ ) or Python with some graphics library such as my MakeBot ( http://stratolab.com/misc/makebot/ ). If people are interested, I can post my Scratch worksheets.
Slashdot Mirror
I've been teaching middle school kids programming for four years. I don't think the language is that important. What's important is to find good problems that they can work on. And by good I mean: (1) Students find them interesting, (2) early problems are easy so everyone can get a small victory under their belt, (3) the problems scale up so everyone has something that challenges them. The way I teach is I start everyone out on Scratch ( scratch.mit.edu ). Scratch uses drag-and-drop programming which lets you skip all the worries about syntax and jump immediately to logic, problem solving, etc. I.e. the interesting stuff. Most of the students will continue to find Scratch interesting and challenging for a long time (6-months to a year). Your top kids will get bored and then I'd move onto something like Processing ( http://processing.org/ ) or Python with some graphics library such as my MakeBot ( http://stratolab.com/misc/makebot/ ). If people are interested, I can post my Scratch worksheets.