This is a series of non-technical watchmes which will give you a pretty good idea about the project, and how you can adapt it to your own experiments :) (Hint: click on Play All to watch them in order)
The watchmes for Part 3 and Part 4, Morphemes in Context, for the Tools for Field Linguistics workshop, April 30 at Concordia University.
Advice: GATE is just one tool in your tool box, you can use grammars, dictionaries, articles to find out previous analyses of the phenomena you're working on. You might be tempted to spend some time putting that information into Gate but that isn't necessary, and its often a waste of time. You should should start with a clean slate, just a Unicode Tokenizer and a research question. You can use GATE to look up just the example in context that you need (it could be a morpheme or a sequence of morphemes or word categories) so that you get more relevant examples early in your research. You also get data that you can check with your informant.
Try to use real data: data from blogs, forums, chats, and subtitles. Check with your informant about which dialect the data is coming from. Often the data can serve as a jumping off point for some clear contexts that will help your informant think of relevant information when they give grammaticality judgements.