Monday I did some relatively intensive last week before Xmas advising of PhD students. Due to my scheduling genius I was talking to four at the same time. I didn't do as bad as I might expect at multi tasking, but I am just not at the same level as my advisors Larry Summers and Larry Katz. In particular Larry Katz totally showed me up on Monday.
I report some e-mail correspondence with considerable editing.
From: Robert Waldmann
To: lkatz@harvard.edu
Date: Dec 19, 2005 3:49 PM
Subject: competing risks
competing risks [snip] and worried a lot about correlation between the effect of unobserved heterogeneity on one and the other. By amazing coincidence I am advising a student (Dario Sciulli) who has exactly the same problem. [snip]
I would guess that Heckman has something with mass points for a general competing risk model. It would be very nice if such a thing has been implemented in STATA. [snip]
ciao
Robert
From: Lawrence F. Katz
To: Robert Waldmann
Date: Dec 19, 2005 3:54 PM
Subject: Re: competing risks
Robert,
[snip] there has been some recent practical work on how to do
implement competing risks models and to allow for correlation across the
risks by Adriana Lleras-Muney and Bo Honore of Princeton. They would be
the people to contact on this issue. A recent versio of their paper is
available from Adriana's web site at
http://www.princeton.edu/~alleras/papers/adbo14.pdf
OK so no big. Larry Katz knows the latest (October 2005) on correlated effects of unobserved heterogeneity in competing risk duration models, but look at the headers.
3:54 pm minus 3:49 pm is five (5) minutes. This is a totally cold e-mail after years with no correspondence and he sent back a URL five minutes after I ask him.
This is a topic on which he hasn't worked in about fifteen (15) years.
Now I made a big mistake. I sent the e-mail and checked for the reply with four (4) advisees watching me. Now they will expect the same kind of service from me, with follow up for twenty (20) years.
Very dumb move.
No comments:
Post a Comment