[Rdap] Deciding when a Data Management Plan is not required?

[Joe Hourcle]

On Thu, 21 Apr 2011, Aletia Morgan wrote:

> 
> Greetings ? I?m working with PIs on reviewing DMP documents that are 
> part of NSF proposals, and I am wondering about whether a DMP is needed.
> 
> The project involves staging a meeting with researchers in the 
> discipline, with a goal that includes the development of plans for 
> future research.
> 
> Obviously, there?s no experimental data here, and the output is 
> essentially a narrative document.  The PI asserts that ?This proposal 
> does not seek to collect environmental data. A data management plan is 
> not required.?

It's observational data, not experimental.  And if you're collecting 
information about the attitudes of people, I'd consider that to be 
environmental, personally.



> If a conference is being organized, notes are taken, documents are being 
> written.   Is there a need to say how these records will be developed 
> and preserved?  Or am I being too compulsive!

I think this goes back to the discussion at RDAP11, where some people 
didn't think of what they were collecting as 'data', but I'd argue that 
the notes taken, and recordings of the meetings, etc, are in fact data. 
If there were some sort of coding done on the participants responses or 
overall attitudes, or results of any polls/voting at the meeting, all of 
those would be more easily recognizable as 'data', but I'd argue that 
notes, even with inherent bias and lack of completeness are in fact a form 
of 'data'.

I'm actually participating in multiple side discussions right now on the 
nature of what 'data' is;  my point of view had been from the digital 
sensor aspect, but in this case it's the recordings by some observer (in 
this case, a person, not a sensor) that are going to be used as the 
supporting basis for any results.

I'd say that the notes on planning the meetings might not be 'data', but 
the agenda and materials given to participants for the meetings are 
important documents to archive, as they're the equivalent of the 
experimental plan -- knowing how you presented information to the 
participants to elicit the responses that you received might either be 
useful to someone trying to repeat the process in another discipline, 
repeat it with a larger audience, or repeat it after time has elapsed to 
see if attitudes have changed.

There might also be something in that plan that could hint at potential 
bias in the results.

...

It may also be that the apparent lack of some sort of hard data is a 
problem in itself.  (I don't know how the specific grant, or what's normal 
for that area, so this might not be the case) ... but if you wanted to 
make sure that there was at least *some* true 'data' as a result of the 
meetings, you could distribute a short questionaire to each of the 
participants (yes, I know, that means getting IRB involved), so there was 
at least some 'data' that came out of the meetings ... even if it's just 
on how well they though they meetings went and comments for how to improve 
them.


> Thanks for any thoughts, and if there might be a better place to ask 
> this question.

If there are any discussion groups for the specific discipline covered by 
the grant (even if that isn't the precise discipline the researchers might 
be in), I'd also ask there.

... but I think this goes into what what Cliff Lynch said were the basic 
parts of the NSF DMPs ... the first step is to figure out what the 
products of the effort are valuable and worth keeping ... and he did say, 
it's possible that there aren't any.

I personally think there are, even if it's just to allow someone else to 
review how you conducted the meetings, so they can repeat them or improve 
on them, and thus save time for the next iteration.  Is it data or 
documentation?  Well, I'm not going to get into that argument, but 
whatever it is, I think it does have value.

-Joe

-----
Joe Hourcle
Programmer/Analyst
Solar Data Analysis Center
Goddard Space Flight Center