I'm a graduating undergrad - but I'm sort of working my way into the Medievalist field. (I'm in the running for an ETA, by the way). I work with 13th century French poetry that treats penitence and sin, looking at authorial voice and the use of shame and guilt.
Just thought I'd add something quickly to the question on the number of recommended candidates. We're all getting this one and a half to twice the number of spots figure, and I think it's coming from here:
http://us.fulbrightonline.org/info_recommended_candidates.html
The key is that at least this number of people are recommended, so while it's masquerading as information, this really tells us nothing.
I don't think that critiquing something is taking it too seriously. Someone offered a method of analysis to a group of people who decide to comment on it. What a rebellious thought.
*Addition*
I think, however, that there are certainly people who wish to pretend that they're slightly less anxious than everyone else to hear notice.
I think there are a few issues with your methods - it appears your regression may have been contaminated by using the notification dates of a country for all categories, instead of separating them - that's my only explanation for why STATA came out with the number of 13 April for France ETA applicants, even though in years past, according to the data, it's always been in March for the ETA's.