Michael Spires wrote:

"The argument I've heard most frequently for deleting or destroying files is that if we keep them after we're no longer legally required to do so, they're still available for audit review and we'd have to produce them if asked, so it's better not to have them around. It seems to me that this argument is presuming that having to produce a file is automatically a bad thing, or that reviewing a file would automatically turn up evidence of wrongdoing (or at least evidence that would result in an audit finding)-but surely if we're all doing what we're supposed to, our files will be in order and easily capable of passing an audit with flying colors. So is there another reason why hanging on to such records would be considered contrary to best practices?"

The argument does not presume that there WILL be a finding, it only presumes that there MIGHT be.  There is no upside to an audit.  Either there's a bad outcome in the form of a finding, or there is effectively no outcome at all.  Why would you do anything to increase the odds, even just slightly, that you'd be on the wrong end?

As you point out, general records of proposals and awards can be very useful down the road, and I don't see any problem with keeping those kinds of things locally.  You are looking to answer questions like, who submitted, what did the science look like, were they funded, what kind of reviews did they get, etc.  That kind of information is perfectly fine for a professor or department to keep for as long as they like.

But that's very different from the central office keeping official copies of hard financial data (who purchased what, when, and using what justification) any longer than is legally required.

-Adam

Adam Kuehn, Assistant Director
Office of Research Support
Duke University
2200 West Main Street, Suite 710
Durham, NC  27705
919-681-8689 (direct)
919-684-3030 (central)
xxxxxx@duke.edu

<br>
======================================================================<br>
 Instructions on how to use the RESADM-L Mailing List, including<br>
 subscription information and a web-searchable archive, are available<br>
 via our web site at http://www.healthresearch.org (click on the<br>
 "LISTSERV" link in the upper right corner)<br>
<br>
 A link directly to helpful tips:  http://tinyurl.com/resadm-l-help<br>
======================================================================<br>