zhanwu and yi

judy cole

christian

---

1. use the existing database to do a simulated sample?

1a. use the existing data base to look at email vs no email differences?

2. advice for building panels and future survey samples?

3. cap/recap analysis of the activities/events/awards/committees [and
   other?] data?  [captures might be time rather than event?] [some
   dependence structure due to depth of commitment?]

mythologies that would be useful to confirm or deny

- differences betwene pre-67 and post-67 (or whatever year) 65-75

- cyert year students vs more recent (roughly late 80's). 85-95

(this is useful for the paneling)