It seems to me that year submitted v year rec'd issue shouldn't
matter too much provided you measure it consistently. You're really
looking for changes over time.

I've been thinking about this general issue for some time, and it
might be that perhaps a more meaningful indicator could be "yield".
Under this concept one would worry less about the numbers of proposals
and "hit rates", and would focus more on something like "cost per $
awarded", or "unit cost per successful grant", "cost as % of average
award amount". A hit rate tells me a narrow part of the "success"
story; a "yield" fills it in a little more. For example, we may both
have 50% hit rates (think big!) but you may be more "successful" if
your grants are bigger, or if it's costing you less to process each
proposal, or you're submitting a lot more proposals, etc. By using or
adding a "yield" indicator you can make more meaning of the hit rate.
I think both "hit rate" and "yield" have the weakness of being subject
to a lot of year to year random variation, especially when dealing
with small numbers of proposals.

Bill Kirby
NSF
703-306-1102

______________________________ Reply Separator _________________________________
Subject: Hit Rate
Author:  Research Administration Discussion Group
<xxxxxx@health.state.ny.us> at NOTE
Date:    7/12/96 10:52 AM

We are a small but growing research institution, and we are trying to
determine the best way to measure our success rate.  We are interested in
hearing how other universities handle this in general.  Do you measure
new proposals submitted in a given year against new proposals received in
that same year regardless of the year submitted?  Do you do a one-to-one
correlation (e.g., if proposal X is submitted to NIH and is funded, that
counts as a hit regardless of the year submitted and received)?

A more specific issue that puzzles us is how should we count a proposal that
is submitted to several different sponsors.  For example, if a researcher
submits the same proposal to three different sponsors, does that count as
3 proposals or just 1?  If the submissions would be counted as just
one proposal, what if the same proposal is submitted to two different
sponsors, but the researcher modifies the name?  Research administration
does not necessarily have the technical expertise to know that they are
the same.

Thanks in advance for any information you may be able to share.

Linda
(xxxxxx@umbc.edu)