Apps/packages that could support the specification

[mekline]

There are a few ways that existing software can work with the specification. In particular, they can read to-spec datasets (reading the dataset cleanly, taking advantage and correctly using the relevant metadata), and can write them (by generating data files and/or metadata that fit it.)

Let's add possible candidates here!

[jodeleeuw]

I will certainly provide an option for jsPsych to output data in a spec-compliant format. I'm not sure if it will be the only output format. I think that would require the existence of some tools that could take spec-compliant data and storing it in various databases.

[mekline]

I doubt we'll be looking for exclusive commitments :) As you said, this is something that gets better if other tools in your ecosystem already use it - which means a chicken-and-egg problem for developers, but I'm hoping that there will be people for whom it's a relatively short jump who can get the ball rolling. So - thank you!

[doomlab]

@mekline - is this where you want us to list the apps we've been working on? I am only half following what's going on because my full attention is surviving this semester and moving across the country, but want to make sure I'm keeping up with you guys.

[mekline]

@doomlab I'm hoping we can collect general ideas here, and then also if there's a repository that people can visit/help out with, it should get its own issue (so it shows up in the list here: https://github.com/mekline/psych-DS/projects.)

(I'm copying this off of some other OS projects in my Mozilla cohort, if this seems like a confusing plan let me know! The idea is to try & keep track of all the activities related to the project in a semi-standardized way so we can keep track of what's going on...)

I'll make both you and @jodeleeuw collaborators on this project so you can add links etc.

[mekline]

(Making an issue for @doomlab 's app to show what I mean...)

[mekline]

@rubenarslan want to add an intro (or just link) to codebook as a new issue?

[doomlab]

@mekline - got it! I am learning a bit of workflow git wise here from you, so keep up the work 👍 . I am also reworking the app we did this summer at SIPS with some things I've learned from writing the data spice one, so you can add that to the list too. It's linked here: https://github.com/debruine/ddcreator

[mekline]

I lost track of this email, sorry! Do you expect ddcreator and the dataspice fork to eventually merge or serve similar purposes, or are they separate? Eventually we should have short blurbs for each package/app, but right now I"m just try to get my head around this :)

…

On Thu, Oct 18, 2018 at 12:17 PM The Doom Lab ***@***.***> wrote: @mekline <https://github.com/mekline> - got it! I am learning a bit of workflow git wise here from you, so keep up the work 👍 . I am also reworking the app we did this summer at SIPS with some things I've learned from writing the data spice one, so you can add that to the list too. It's linked here: https://github.com/debruine/ddcreator — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#10 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AGgP9Vy9v2TcHVgapVNRGduUfcRSKLuuks5umKmdgaJpZM4XEpM1> .

[doomlab]

@mekline

At the moment they seem to do different things. I think we could merge them ... may take some pilot testing to think about what's easiest and what fits best with the spec. This week I am updating all the instructions and the students are making how to guides for the paper we are writing (for updated dataspice). Mainly I see it as:

dataspice - fits nicely with scheme.org
ddcreator - does more for labeling and sharing data coherently

But I may have a better answer after working more on both them. We will have tutorials for all the things by the end of the semester that we can share! :)

[schnarrd]

Leaving this thought here: It would be wonderful to work with OSF to integrate the validator app into the OSF data storage options. What I'm thinking is one of the component options could be something Psych-DS-Dataset which would be displayed where the red box is:

In order to save the dataset on OSF, the dataset would have to pass a validator check. This would help immensely with project compliance & visibility.

This is more of a longer-term idea once a validator is programmed and beta-tested.

[FelixHenninger]

Hej Melissa, hi everyone,

a very happy new year, and thanks for all of your work on this awesome project -- I love what you're doing! Ever since we chatted at Mozfest, I've had support for Psych-DS on my todo list -- sorry for the relative radio silence (I only recently noticed that I had a task in your November newsletter). I recently had some free hacking time procrastinated a bit, and added automated sidecar template generation to lab.js. It's not perfect, in that we still need to pick up user-defined data columns, but it's a start.

Here's what it looks like:

I'd love to hear your thoughts, especially around how we can support metadata from the data generation side (i.e. beyond supplying the dataset_description.json). @doomlab and @rubenarslan, I imagine that it would be super-cool if your tools could also import half-baked dataset descriptions like ours (that is, if they don't already), allowing users to correct and extend automatically generated metadata. @jodeleeuw, please do let me know if I can support you with the jsPsych integration, I'd be happy to share our (pretty basic) code.

I also left a couple of comments in the specification which I'd be happy to clarify and discuss if that's useful (please don't hesitate to ping me via mail/Slack, I'm fighting an onslaught of email right now and might miss automated notifications).

Thanks again for getting me on board, I'd love to help with this! Take care, -Felix