Diagrams and "looking ahead" sections

[luizaandrade]

Adding diagrams resolves #461. I went back and forth between having them at the end of the chapter or in the beginning, but it seemed to me that it could go well with the final chapter paragraphs mentioned in #482, so I left them at the end. Let me know what you all think.

[bbdaniels]

These are so awesome! Very minor things left:

• [ ] Overall
  • [ ] I think the green and the blue in Design and Acquisition are really close but I'm OK with it. Very calming color palette choices 😌
• [x] "Intro"
  • [x] Change file name to "Introduction"
• [x] "Acquisition"
  • [x] Center text on "Manuals and protocols"
• [x] "Cleaning"
  • [x] Typo "Recodie" should be "Recode"
• [x] "Publication"
  • [x] Outer border color and box fill does not match header color, it should be the color from "Conclusion"
  • [x] Change "Manuscript" to "Final output"?
  • [x] Not sure I love the overlapping icons but happy to hear what everyone thinks
• [x] "Conclusion"
  • [x] Should "Data Work Map" just be "Data map" (watch capitalization)
  • [x] Should "Cleaned data" be "Clean data"

[kbjarkefur]

First of all, these diagrams is sooo good!

General comments

• [x] I think we should use only one icon for documentation, and not the "multiple documentation icon" used, for example, in he Acquisition diagram for Manuals and Protocols. Many of the documentation that has the "single documentation icon" will likely also have multiple components. I do not think the point of indicating what is more than one thing justifies not making it as simple as possible.
• [x] I think it should always be "disclosure risk" and not "risk of disclosure". Disclosure means making something known. We want to make things known when we publish things. But when we disclose things there is a risk that we disclose things we do not want to make known. So there is a risk associated with disclosure, ie. "disclosure risk", but some disclosure is often intentional and therefore we do not worry about "risk of disclosure". Both in intro.png and publication.png

Cleaning.png

• [x] The end point is called "clean data" but it is called "cleaned data" in the conclusion, processing and publication diagram. We should be consistent.

Analysis.png

• [x] I am not fan of the curly bracket as it brakes the style. Can we do three regular arrows from the three boxes "Treat missing...". "Integrate data..." and "Create analysis..."
• [x] If we should be consistent then there should be an arrow from "Analysis data" and "Analysis code" to "publication"
• [x] Polish the results. doesnt it sound as if we try to make the resutls better than they are? Isnt "formatting results" better?

publication.png

• [x] Why do we make a difference between "License and terms of use" of replication package and "terms of use" for data publication package? Why no license for data?

intro.png

• [x] Why is only non-original data have an arrow back to sampling? I do not think we need this arrow, but if we should keep it, isnt it more fitting that the original data branch point to sampling?

conclusion

• [x] we do not mention original data in any other diagram

[luizaandrade]

Analysis.png

• [x] Polish the results. doesnt it sound as if we try to make the resutls better than they are? Isnt "formatting results" better?

Does "polish and export outputs" sound better?

publication.png

• [x] Why do we make a difference between "License and terms of use" of replication package and "terms of use" for data publication package? Why no license for data?

We typically release terms of use for the data on the microdata catalog, and add a license to the code on GitHub. But if that doesn't seem right, I can change it.

[bbdaniels]

@kbjarkefur are you done working here?

[kbjarkefur]

Yes, but someone should review my last commit

[bbdaniels]

Done, although I thought the captions didn't fit and removed them. The intro one is fine, let's see what other people say before fiddling any more with placement. Merging!