Unable to connect figshare account to OSF without institutional login: solution?

58 views
Skip to first unread message

Florian Dierickx

unread,
Jun 7, 2019, 9:18:55 AM6/7/19
to Open Science Framework
Dear all,

I am very enthusiast about the OSF framework and trying to get background data (figures, data sets, ...) of my thesis organized on the platform. To have a properly referenced and organized set of data and figures, I use Figshare to reference the different parts.

Now, when I try to connect the Figshare data to the OSF framework as described in the OSF FAQ Guidelines on establishing a Figshare connection, I don't manage to get step 2 (Configure Figshare) working. Every time I tried to connect the data, the connection link points me towards a login from an institutional account (both with Safari and Firefox on OSX).

--> Is this normal behavior and a prerequisite to be able to connect? If not, I would be happy to read and learn a solution!

Best regards,

Florian

David Mellor

unread,
Jun 7, 2019, 9:58:49 AM6/7/19
to Open Science Framework
Hello Florian,

Thanks for reaching out. There may be something odd going on between the two accounts. I've tested on my own personal account and it looks to be working as expected so maybe there's an institutional account setting either on Figshare or OSF that is causing an unexpected behavior. Can you submit a request here and we'll work to figure out what's going on: https://openscience.zendesk.com/hc/en-us/requests/new  If you attach a screenshot of the problem to that request it could be helpful, too. 

Thanks! 
David

Sara Bowman

unread,
Jun 7, 2019, 10:05:57 AM6/7/19
to Open Science Framework
Hi Florian,
Thanks for getting in touch. I'm not able to reproduce the behavior you're seeing - can you send an email to sup...@osf.io and include a screenshot? Our support team can help troubleshoot a little more in depth.
Thanks!
Sara

Florian Dierickx

unread,
Jun 7, 2019, 10:32:36 AM6/7/19
to openscienc...@googlegroups.com, sup...@osf.io
Before going into depth in the replies below: I tried the same pathway with the Chrome Browser [75.0.3770.80 (Official Build) (64-bit)] and the figshare connection works as expected !
For the sake of completeness (and in case you want help or testing for what seems to be a Firefox or Safari bug), I remain available for testing if you deem it necessary for the platform.

---

Dear David Mellor and Sara Bowman,

Thanks a lot for you fast and extensive response! I submitted a request through the https://openscience.zendesk.com/hc/en-us/requests/new link, as well as on GitHub (https://github.com/CenterForOpenScience/osf.io/issues/9059)

Dear OSF Support Team,

As suggested by Sara, (in case not doubling the request through the zendesk / github links) hereby a request for troubleshooting : when I try to connect the Figshare data to the OSF framework as described in the OSF FAQ Guidelines on establishing a Figshare connection with Safari or Firefox on OSX, I don't manage to get step 2 (Configure Figshare) working. Every time I tried to connect the data, the connection link points me towards a login from an institutional account. I am using Safari [12.1 (14607.1.40.1.4)] and Firefox Quantum [67.0 (64-bit)] in macOS Mojave [10.14.4 (18E226)] , with adblockers (uBlock origin) activated and deactivated. Please find also a screenshot attached

It seems the redirection pathway after clicking on 'Connect Account' is:

--
You received this message because you are subscribed to the Google Groups "Open Science Framework" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openscienceframe...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openscienceframework/dd252e16-817a-4bc0-a3b2-dd7481a0992a%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


--
Florian Dierickx
EU ITN Marie-Curie PhD Candidate
Hybrid IO/LCA Modelling of Renewable Energy Deployment
CERDI UCA-CNRS & Iceland University
+33 641 00 45 37
skype florian.dierickx
twitter FlorianDRX
linkedin floriandierickx

Stanka Fitneva

unread,
Jan 27, 2020, 2:58:48 PM1/27/20
to Open Science Framework
Hi all,

We have a dataset that we would like to test some additional hypotheses on. They are unrelated to the original question, so the sample seems to be almost 3 times what we need! (Don't see this problem often!)

What would be your recommendation how to proceed with analyses and preregistration?

I see a couple of directions:
  • split the sample into 3, run the planned analyses on all 3 and see how the effect sizes align
  • draw one sample to run the planned analyses; use the rest for confirmatory tests (the study design is not an exact replication of the study on which we are basing effect size estimates). In this case, should we preregister just the first analyses or both?
  • use the entire sample noting that study is over-powered for tests
Opinions / suggestions would be much appreciated!

Thanks!
Stanka

Stanka A. Fitneva, PhD

Department of Psychology

Queen's University

Kingston, ON K7L 3N6

 

phone: 613-533-2363



David Mellor

unread,
Jan 27, 2020, 3:39:55 PM1/27/20
to openscienc...@googlegroups.com
Hi Stanka,

This is a good problem. My recommendation is a slight modification your first and second plan. You already have a preregistration, correct? If so:
  1. Run the planned, confirmatory analyses on a random third of your data. 
  2. Then dig around a lot in a very exploratory way. 
  3. Preregister a 2nd time once you have anything else worth looking into. 
  4. With the second third of your dataset, run those new analyses (you can ALSO run the original analyses and do some direct comparisons of effect sizes, as your first item suggested)
  5. Repeat 2-4 if you have a reason to do so. 
Another option would be to do steps 1 and 2 and look into using the latter two thirds of the data as the basis for submitting as a registered report (cos.io/rr).

Good luck!
David

--
You received this message because you are subscribed to the Google Groups "Open Science Framework" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openscienceframe...@googlegroups.com.
Reply all
Reply to author
Forward
0 new messages