[open-science] Open Science at the British Ecological Society annual meeting

Thu Sep 15 15:24:43 UTC 2011

Well,  if one plans to do further analysis with some data, one should
reserve a time window , say one year? two years? and then do it
To sit on a dataset indefinitely and then after one year, two years
nothing has come out, sounds wrong

What about an option to preserve confidentiality of the data, provided
analysis is published on a quarterly basis. If and when publication of
the results does not occur, the dataset should be opened to others
(nice incentive to publish:-)

Other obvious reasons for people not wanting to share their data. The
first that comes to mind is (pardon the cynicism) their data is
'no good', therefore not presentable. The general public may never
know (and science administrators may have forgotten) how messy things
are in the lab. Or even worse (double forgive pls) The data does not
exist, or it has been completely fabricated. Therefore cannot be
shared. I dont know any reputable scientist, unless they have a patent
in mind, who wouldnt be happy to open the doors to their science,
likewise, I know many fabricators of science who pose as scientists,
who are horrified at the thought of someone looking into their
business

Other considerations
- its the way that science is done (competitively) that makes one
suspicious of opening up their data. Scientific journals compete for
original results, and research funding bodies instigate this
'competition' by means of rewarding 'exclusivity' of scientific
knowledge.  So
if people could get, say, more funding, for making their datasets
available, surely they would do it.

-  a few more rules perhaps? for example, when publishing results, one
must also publish the underlying data upon which the results and
conclusions are drawn, otherwise its not science, but fiction

- when not publishing datasets for everyone to use, one should come up
with further analyses within certain timeframes, not just hoard

- journals take ages, eons, to publish research results. not
necessarily because it takes time to 'check' the data. sometimes is
just red tape and holding back new knowledge. Reputable journals could
publish 'submitted papers' for everyone to see and openly review, then
also publish the 'accepted for publication' ones, based on their own
peer reviews, which would also make the peer review more transparent
and accountable.

Bottom line: its the whole system  that needs to change
not just 'access to knowledge' but also how scientific knowledge is
generated and produced, funded, disseminated, validated, reviewed,
replicated etc

(not expecting personal popularity rating to go up with these
suggestions. just to instill some ideas...)

...probably more.....

PDM

On Thu, Sep 15, 2011 at 3:58 PM, Thomas Kluyver <takowl at gmail.com> wrote:
> On 15 September 2011 15:25, Ross Mounce <ross.mounce at gmail.com> wrote:
>>
>> The key to this is valuing your data being re-used by someone else
>> (data citation).
>> At the moment people see this re-use as bad [scooping], rather than a
>> good thing [my data was re-used, I got a citation for this re-use, win
>> for me, win for them, win for science]
>
> I think this depends heavily on the nature of the reuse. For meta-analyses
> like those you're doing, I expect people wouldn't have a problem with
> contributing their dataset, because it's something they're probably not
> planning to do themselves. What I think worries people is that someone else
> could just pick up their dataset and do the same analyses they want to do.
> Also, I think that even people who're happy to share data may feel more
> comfortable if re-users have to get in touch with them and request the data,
> than if it can just be downloaded freely.
>
> I'd love to see better ways to give credit to people assembling good
> datasets (not least because that's what I'm currently trying to do). But I
> don't think that's the whole story. Data collection tends to be slow and
> tedious, and regardless of the career implications, scientists don't want to
> be the one doing the hard graft of collecting data for someone else to do
> the relatively interesting work of drawing conclusions. Perhaps ecology is
> particularly affected by this - data collection can be particularly slow,
> and the statistics are often very complex, which might lead to fears that
> someone who knows the stats well will overtake the first researcher on the
> analysis.
>
> Thomas
>
> _______________________________________________
> open-science mailing list
> open-science at lists.okfn.org
> http://lists.okfn.org/mailman/listinfo/open-science
>
>