Lab Managers-
(A long summary seems to be an oxymoron, but there were a lot of
responses, this is a summary of those responses, but it is still a long
email message....)
You may recall that a couple of weeks ago I raised the question of
how one allocates time when the needs of two groups are greatly different-
a couple of hours per day for short term experiments, and 24-36 hour
blocks for long term experiments. Here is the promised summary of your
responses and wisdom. I would like to thank the 16 who responded:
Woody Conover, Paul Cope, Phil Dennison, Ron Federspiel, Jack Fowble,
Charlie Fry, Neil Jacobsen, Joyce James, Peter Lundberg, Rudi Nunlist,
Stephanie Mabry, Dave Scott, Jeff Simpson, John Snyder, Thomas Stringfellow,
and Wei Wycoff.
The quick summary bottom line is that there is no good, easy solution
to this problem, and that it has the potential for being a political
minefield. Several warned of the dangers of getting wedged sideways between
2 or more groups without a GOOD set of guidelines for allocating time.
There were two 'outlier' points: duelling pistols and another
instrument. While another instrument may be the ultimate solution, it will
not happen in a useful time scale. Duelling pistols may work in a more
timely manner but would probably have an adverse effect on facility funding.
There were numerous suggestions for schemes to break up the week into
long blocks of time and shorter blocks of time, and it seems clear that
we will have to adopt some form of this. One interesting suggestion was
the addition of an oversight committee that would review time requests to
attempt to make the time division more equitable. This has the advantage
of removing such decisions from the lab staff, but would still require some
sort of well defined basis for making decisions, as well as a group willing
to do so on a regular basis. I'm not sure how workable that might be in
practice.
Another suggestion, that might also be useful, is to make long blocks of
time available first, maybe 5 or 6 days in advance, and shorter blocks of
time available on shorter notice. This coincides well with how people will
normally want to schedule time anyway - you generally know well in advance
when you are going to want a longer block of time. This system clearly
favors those wanting long blocks of time, however, and that is where my
problem resides. Those wanting long blocks can probably completely fill
the week.
Another interesting suggestion was to pre-allocate time for the different
users, but to do so not only on the basis of their requests, but also on
the basis of how efficiently they used their time in the previous allocation
period. This effectively penalizes those wasting spectrometer time and
optimizes the amount of work done on the system.
Another suggestion was to place limits on the time reserved both by user
and by group, so that not only would a user have a limit on the amount of
time that could be reserved, but the group as a whole would as well.
One manager suggested not only different size blocks between daytime,
evening, overnight, and weekend time, but also a differential rate structure
to make the less desirable times more desirable. In our case, there is no
problem getting people to use the overnight hours, so this is not really a
solution for the current problem here.
The most innovative solution, assuming that the long blocks of time were
for samples withl long T1 values, was to time share the magnet, with the
long T1 samples kept polarized in the magnet's fringe field. The mechanical
implementation of this seems difficult, and in any case this would not
address the problem here. The long blocks of time are being used for
high resolution, high signal to noise spectra with quantitative integrations.
The idea that came across most clearly was that this is really a problem
that the users need to resolve cooperatively. Any attempt by the lab staff
to impose a solution without the cooperation of the users themselves is
almost guaranteed to fail.
Again, let me thank all of you who responded to my plea for help. If
nothing else, I am more confident that I have not missed a simple, obvious
solution, and I can work at getting the users to define a better cooperative
structure so that none of them suffer unduly, and all of them get their
work done in a fair and fairly timely manner.
Thanks again for all of the responses.
Steve
-----
Steven K. Silber E-mail: s-silber_at_tamu.edu
Texas A&M University voice: 979-845-1745
Department of Chemistry fax: 979-845-4719
P.O. Box 30012 URL: http://www.chem.tamu.edu/services/NMR/
College Station, TX 77842-3012
Received on Mon Mar 25 2002 - 18:24:00 MST