[open-science] Let us denonce the pseudo-open Public Library of Science

Thomas Kluyver takowl at gmail.com
Tue Feb 14 16:08:03 UTC 2017

On 14 February 2017 at 15:54, Heather Morrison <Heather.Morrison at uottawa.ca>

> My argument (so far) is that replicability is a poor argument for open
> data. Open data does not facilitate replication, nor is it necessary to
> replication.

I think we may be debating the meaning of the word 'replication', but as I
see it, open data facilitates *partial* replication. We shouldn't pretend
that rerunning the computation totally verifies a result, but it's not
pointless either.

I believe that most researchers would never fabricate data, but that it's
quite common to tweak the analysis a bit to get an interesting result that
you can publish. If we have the raw data and the computational steps done
on them, that's the starting point for evaluating: Does it hold up if I use
this kind of analysis? What if I change this assumption? That point looks
like an outlier, what happens if I exclude it?

That doesn't mean we should have 100% confidence in the data going into the
analysis. But I think it would be a big step forwards for reproducible
research if it was normal for the analysis steps to be reproduced.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.okfn.org/pipermail/open-science/attachments/20170214/daa62a45/attachment-0003.html>

More information about the open-science mailing list