Your proposal should clearly answer two questions: what you want to do and why we should care. Please don't spend too much time on a literature review unless it is needed to understand what you want to do and why we should care. This is a chance for you to get supportive, helpful feedback on early stage ideas so you should feel free to take some chances and be open about the weaknesses in your proposals.
Each proposal is limited to 800 words. As you will see, 800 words is quite short so use these words wisely. You are encouraged to include figures and graphs to illustrate your ideas, and these do not count against your word limit. All proposals must have clear titles. Proposals must be submitted Sunday at midnight.
You will spend a lot of your career giving feedback on people's ideas. This is an art, and this is your chance to practice that art. The goal of your feedback is to be helpful to the author; it is not make yourself look smart.
In writing your feedback, it is useful to start by summarizing the goals of the proposal as clearly and succinctly as possible. Then, you can describe things that you liked about the proposal. Finally, you can make suggestions about how the research described in the proposal can be improved; the more concrete and specific your suggestions the better.
Proposal review must be done between Sunday at midnight (when the proposal is posted) and Monday at midnight.
In order to help you learn more about how to give feedback, the author of the proposal will give you feedback on your feedback, such as which parts she found most helpful and what changes she plans to make based on the feedback.
The final proposal is designed to get to take the first step from proposing research to doing research. Thus, in the final proposal you must 1) propose a project (it could be one of your earlier proposals) and 2) take one concrete step toward actually doing the research. This step could be things like acquiring and summarize some data; running a pilot test of your survey on your friends; doing an experiment on Amazon Mechanical Turk; or having some friends do some labeling of images and then trying to learn from it. The goal here is not to do something perfect; it is to do something.
Your final proposal can directly build on one of your earlier proposals; in fact, that's what I would recommend. But, if your proposal is based on an earlier proposal, please highlight which parts are new. The proposals were limited to 800 words, and the final proposal is limited to 1,200 words.
You are encouraged to collaborate on this final projects with someone from outside of your field. One way that you could find collaborators is to think about earlier proposals that you really thought were interesting.