Flow of formatting errors/metadata section (episdoes 2-3)

[brownsarahm]

The organization of the beginning of this lesson as written feels either repetitive or mysterious. I taught it for the second time today and I switched the order some that made it flow a little more smoothly I think. I'd like to recommend a fairly significant reorganization. of the content in the first 3 episodes.

Introduction [new ep 1]

( material that's there already)

Data Description

overview of the SAFI data (from ep 2)

Formatting Data in Spreadsheets [new ep 2]

intro with goals and basics of tidy data (from ep 2)
exercise to find errors in messy data (from ep 2)

(common formatting errors sections from ep 3)

MetaData [new ep 3]

current material on metadata (from ep 2)
new material on how to setup meta data// links to resources,

the suggested new material came up because the host site I taught at today the librarians in the room (hosts) mentioned that some repositories have standards for metadata and standards for it, we don't need a deep coverage of that, but links or points to some of that and noting that format of meta data is based on disciplinary standards I think is valuable add to make that discussion more concrete.

[chris-prener]

Really appreciate the feedback @brownsarahm - @ErinBecker - since this is a significant change, we should probably kick it back to the CAC?

[ErinBecker]

@chris-prener - totally ok to re-structure and add these resources that @brownsarahm suggested without passing up to the CAC. The CAC did discuss this section at their last meeting (several months ago), and agreed that it was important to have a section on metadata, but didn't have strong opinions on what should be included in that section or what the organization of that section should be. At this point @brownsarahm has taught these materials more than anyone else, so I'm happy to go with whatever she proposes!

[chris-prener]

OK sounds good - I'll be able to dive into this next week I think!

[brownsarahm]

I can also at least start this and I've sent a note to the host from NYU to ask for more resources on metadata.

[chris-prener]

I'd love the help @brownsarahm - do you want to take a first pass and then I'll take a second?

[VickyRampin]

Hi all -- I am the host at NYU that Sarah mentioned above 😄 During the workshop we were discussing codebooks and data documentation, and I mentioned that for folks who don't know where to start, some repositories have samples that are useful (with the caveat that these samples are often meant to demonstrate not only detail but required formatting which I tell people, unless you want to submit there, don't worry about the formatting excessively).

For instance, here's a sample codebook from ICPSR, one of the most popular repositories for social science data: https://www.icpsr.umich.edu/icpsrweb/ICPSR/help/cb9721.jsp

I might give this to a social science researcher who has never documented their data before, and say something like "see how they put their variables in a table saying what they are? that's a good first step"

Some other good examples/help guides:

• https://dss.princeton.edu/online_help/analysis/codebook.htm
• https://www.ukdataservice.ac.uk/manage-data/document/data-level/tabular
• https://www.ukdataservice.ac.uk/manage-data/document/data-level/qualitative

The other piece I mentioned in the workshop that may or may not be helpful is that there are different types of metadata -- in this case, descriptive metadata and discovery metadata. Descriptive metadata is supposed to describe the content of the data (e.g. codebook) with some administrative metadata (e.g. who created or derived the data, with what software, on what date). Discovery metadata is what people search on, more useful when publishing data in a repository (e.g. when you upload something to the OSF or figshare, it asks you to enter some information -- that's discovery metadata).

I can try to work on these materials with @brownsarahm, though this week is really tight for me. Hopefully these resources help get you started!

[chris-prener]

this is great @VickySteeves - I really appreciate the detailed notes. I really like the ICPSR example and your example for searching OSF as an example of discovery metadata. Up to you how you and @brownsarahm want to organize next steps. I'm happy to help, or take a second pass - totally up to you both!

[brownsarahm]

I started a PR #76, allowed edits from maintainers and added @VickySteeves as a collaborator so we can keep iterating on it.

[chris-prener]

Great thanks @brownsarahm - I'll try and check in on this next week!