December updates and dataset contribution instructions!

[Melissa Kline]

Hi all - some exciting developments in Psych-DS land! As usual, here’s the Cliff Notes, and details below

- Melissa

*******

(1) Time to upload datasets! Sign up by December 20, aim to complete to-spec dataset versions by Jan 15.

(2) Hello to Dataspice :) (Erin, Bryce, Anna chime in please!)

(3) Mozilla Open Leaders presentation (Tomorrow/Tuesday at 

(4) Looking ahead to the validator - please help me plan!

*******

(1) We have ‘frozen’ the specification at v 0.1.0 so that we can take a stab at actually applying it to some real datasets and find out what happens! If you wrote to me before about contributing data, this is your moment! You can also add any dataset you like so long as you have permission to share that data. If you have data you can share but *can't* take responsibility for bringing up to spec, or if you'd like to help but don't have data to share, let us know and we can do some matchmaking!

More instructions here - to help out our future coding efforts, it is encouraged to contribute to the Github repository directly, but there is also a parent project on OSF which will allow you to add a folder in a drag-and-drop format. 

For this project, more, smaller datasets are more helpful than fewer, larger datasets - we’d like to feature a variety of topics *and* tools: ideally, we’ll end up with instructions for people who use many different workflows.

(2) Welcome to the makers of dataspice, a series of R scripts that implement the Schema.org Dataset metadata (which Psych-DS also uses)! Erin Buchanan has been working along on a version, and Bryce and Anna from dataspice are excited to coordinate with the Psych-DS project! I’ll let the three of them chime in with next steps?

(3) I am wrapping up the Mozilla Open Leaders training with a brief presentation on Psych-DS tomorrow. I’ll be on at about 12pm EST and you can find the livestream at this link.   You’ll all be extensively thanked; this project has been really rewarding so far and it’s entirely thanks to having such an enthusiastic & active community to work with! 

(4) We are moving at a pretty steady clip toward being able to launch (!) this specification; once we have a set of sample datasets (and the resulting modifications to the specification), a major remaining task will be to create a validator app - rather than just producing a dataset with e.g. an R package or by hand, this app takes a folder (however it was produced), and checks *whether it meets all the requirements* of Psych-DS. This is a critical part of the project, because the validator will in some sense be the ground true of the specification - most researchers will (hopefully) not have to read the document, but will use some version of the validator to check whether they’ve succeeded, and some app creators will hopefully be able to draw on this code directly as well. 

I’m hoping that we can start from the BIDS validator (forked here), since it has many shared features we’ll ultimately want; though since we meet the Schema.org Dataset specification there may be other tools we can draw on as well. 

Challenge - Melissa doesn’t know how to build this on her own! In particular, I am not yet sure whether this community has people with the skills *and* time/willingness to make the adaptations, or whether we need to seek more help. 

If you might be interested in this, OR if you are a person familiar with the BIDS validator and can say more about what will be involved in implementing these changes, please speak up! You won’t be committing to doing anything at this stage; I think we’ll hold off on this until we have some actual datasets to validate, but I want to get a sense now so we can figure out whether or not we need to be looking for more outside help.

Thanks all, and I hope your end-of-semester and holiday busy-ness is treating you well!

- Melissa 

-----
Melissa Kline
Postdoctoral Fellow
Harvard Psychology/MIT Brain & Cognitive Sciences

http://www.melissakline.net
mek...@mit.edu

I check email about 2-3 times a day. If you need something from me in the next 4 hours please call or text me on my cell phone!

[Anna Krystalli]

Hello everybody, and thanks for the invite Melissa!

We're excited to be able to join forces with this project and integrate the excellent work Erin did, building on the dataspice code. For anyone interested in following along the discussion, feel free to drop by the dataspice repo, and in particular, this issue where we'll be mapping out next steps.

Cheers

Anna

[Erin Buchanan]

Heyo! I was knee deep in moving during this time - I'm happy to still provide a dataset if you need that! Can also work on the validator potentially depending on how you envision that happening.

erin