Group proposal

[andrewpbray]

Please push your proposal to this repo.

[joeyalowitz]

Pushed!

[andrewpbray]

Apologies for the delay on my end.

1. This would strikes me as a good area to work in generally, but predicting locations really should rely upon a model that explicitly incorporates the spatial correlation. Since we don't cover those at all in this course, I'd steer you away from this.

2. Interesting topic, but I'm worried that you'll not have enough data to do anything that is methodologically interesting (only ~50 countries in Africa, right?).

3. This one is the most promising. My only reservation is how you will collect enough data to make this interesting. What is your plan for recruiting people to fill out your survey? How many questions would you plan on asking? You might also see if you can find data sets that already exist out there that have this info. There's one on pottermore that I just came across . . . but I can't tell if the response was selected by the respondents or by generated by the quiz itself. If the latter, it's not very useful.

Please get back to me as soon as you can about these.