Maybe I can help directly, I've done such a thing a view weeks ago.
I simply started the browser by the custom-item-type and presented a html-page within the browser with the task description (you have to make one and put the location of the file in the parameters). By using the <form>-tag and a little javascript which you put in the header of the html-file you can validate participant input. If you use a value that the participant isn't able to respond with, everytime a windows pops up with a message you want (e.g. "NO! Try it once again!", you have to write the message in the code of the javascript) and a simple "OK"-Button.
I added a line in the task description that they can continue if they close the browser window. If the participant do that, the browser window is closed (of course) and ML tells the participant to click on "Continue". Just click "WriteRT" in the ML-file of the custom item from wich you call the browser and you also have a measure how hard the participant tried to answer the question (the RT of the custom item type, or to put in another way, how long they stay doing the custom item).
If you are at least a little bit firm with html-programming, this works really nicely.
BTW: I use Openoffice to create the whole "workaround" of the HTML-file (e.g. tables and stuff like that), because it works like a WYSIWYG-Editor and also provides relatively "simple" hmtl-code (don't try to do that with MS-Word, it's horrible ;-)). So you just need a little bit of understanding of the code, but not extraordinary html-programming skills. In a second step I use a html-editor (in my case Aptana Studio which you also get for free) to modify the html-code (e.g. to put in the javascript).
Best,
7om




Reply With Quote