dime-data-handbook
dime-data-handbook copied to clipboard
One master final review issue
One final review issue to rule them all. Feel free to add more things to this issue, but please follow this organzition:
- Each thing should start with a checkbox
- [ ]
so we can track what has been done - Each comment box is a topic, add things by editing the appropriate box (if you can) instead of create a new comment box. If you cannot edit, just create a new comment box and I will fix
Consistent writing
- [x] Write dataset not data set (old #413)
- [x] Write original data not primary data or raw data when referring to the dataset the project acquires. If talking about data collection specifically you can still say primary data and raw data (old #387)
- [x] Write analysis data not constructed data or analytical data (old #475)
- [ ] Write confidential data not personally-identified data (or PII data) when talking about what data is sensitive and needs to be encrypted. PII data is one reason for data being confidential, and in that specific case we can say PII data (old #454)
- [x] Write flowchart and not flow chart
- [ ] Check usage of "ID", "Project ID" "identifier" and "identifying variable" across the book.
- direct identifier - anything that directly reveals the identity of the observation, for example PII
- identifying information - direct identifiers, but also information that is indirectly identifying
- ID a number or string code for an observation, this should never be a direct identifier
- Project ID the main ID used of a unit of observation in the project. There may only be one per unit of observations in each project.
- identifier - either ID or identifying information
Book structure
- [x] review the sections and paragraphs follow this rules (old #423)
- each paragraph with a statement followed by supporting evidence or examples
- each subsection with a structure
- make sure to draw the distinction between what others have done and what our contribution has been.
- [ ] make sure we call sub-section for sub-section and not section (old #455)
Chapter diagrams
- [ ] make sure font size is good when diagrams are in book (old #474)
- [ ] add an appendix with a full time line of a project. See how that fits in with diagrams (old #417)
Links in book
- [x] Check that URLs that only point to allowed resources. (old #428 ) Everything else should be moved to Wiki, defined in sidenote, or cited as an academic reference. Allowed URLs are:
- DIME wiki
- Any worldbank.org link (including blogs, docs, etc.)
- Any WB/DIME GitHub repository
- Links to top-level domain of specific sites or services referenced
Make sure concepts are explained
Concepts used in each chapter:
(Partly from old #420)
Chapter Intro and 1
- [ ] Reproducibility
- [ ] Transparency
- [ ] Credibility
- [ ] Computational reproducibility
- [ ] Master script
- [ ] Pre-registration
- [ ] Pre-analysis plan
- [ ] Registered reports
- [ ] File-drawer problem
- [ ] PII
- [ ] Data ownership
- [ ] "Confidential data"
- [ ] IRB
- [ ] Encryption
- [ ] Human subjects
- [ ] De-identification
- [ ] Anonymization
- [ ] Risk of disclosure
Chapter 2
- [ ] plaintext files
Chapter 4
- [ ] pseudocode
Make sure all concepts are indexed correctly
- [ ] Make sure indexes that should be indexed are indeed indexed
- [ ] Agree on capitalization and use consistently
- [ ] Combine duplicates or very similar concepts
Indexing:
It is \index{} and \sidenote{} in conjunction, and sometimes new line in text editor. To solve this try the following:
Since \index{} does not appear in the text, see if it can be moved so it is one word, or even just a full stop removed from \sidenote{}. Obviously do not move it too far as we do not want the page references on the index page to be imprecise. \index{} should not be on a new row by itself. Make sure that some word or character follows on the same line as \sidenote{}, as in the closing bracket } in \sidenote{} is not the last character on that line in the text editor.