Moving a discussion from PM with @jd-foster on julia discourse to here. We seem to be in agreement that the actual domain is less important than the specific learning objectives. I wrote (partly paraphrasing @jd-foster in places)

In very general terms (and this is somewhat reiterating what you already said), I think

• IO for a standard data type like json or csv, and io for some other non-standard data
• Organizing data in tabular (DataFrame) form, and as arrays or Dicts. In my field, I often use some custom structs, though I don’t know how universal that is
• Some basic stats and descriptions of data
• Some basic visualisations

Here's a first draft of formalizing these into concrete (though high-level) learning objectives:

After completing this course, students will be able to...

1. Utilize existing julia libraries to read and write files in standard data formats (eg. CSV and JSON) into suitable data structures
2. Make use of general I/O and string-manipulation utilities to read data in non-standard formats into basic data structures such as arrays and dictionaries.
3. View, describe, and manipulate numerical and text data in a tabular format using DataFrames.jl
4. Calculate statistical summaries of numerical data
5. Generate visual representations of data using an existing plotting library.

While lower in priority to #1 or #2, we should decide on the software mode of delivery for the lesson. Each complete SC or DC lesson requires some setup from the learner in terms of interface (GUI or otherwise), browser or manager to facilitate the lesson ideas. This is should not be (but sometimes is) a major barrier to entry for the learner, so the advantages or drawbacks of the software required should be weighed carefully against getting started easily and eventually enabling good practices.

Here are a few options to consider:

1. Julia REPL via terminal: most direct and immediate start, can be intimidating as "just a prompt".
2. Jupyter notebook: harder to setup, but enables better learner feedback loop in process. Adds another layer of "cognitive load" to get through for new learner, learning Jupyter interface and Julia at the same time.
3. VS Code with Julia extension: again, much more to setup than 1 or 2 initially but may pay off in terms automation of code editing, documentation and plot integration. Might be more natural to those used to a browser/app mode of interface.

Any I've missed?

Though (in my opinion) less important than #1, we nevertheless need to identify what knowledge domain we will use as a backbone for the lesson. In PM, I wrote:

I think we should spend the bulk of our energy on solidifying what we want to teach, the actual domain is almost irrelevant. That said, I think features of the data should be:

• Accessible, by which i mean almost everyone can understand what the data is with minimal context. Even if it’s domain-specific, some things are more understandable to outsiders than others
• Inclusive, which is related to (1), but distinct. Eg. Everyone can understand wins/losses of sports teams, but not everyone is into sports. We should try to find something that had broad appeal
• Evergreen - one thought I had was that doing something with corona virus data would be cool, but this may feel a bit less engaging 5 years from now

Despite that last point, a coronavirus - based project could definitely fulfill the first 2, especially since it could include both epidemiology as well as biological (sequencing) data types. I'm a little biased here though, as this is my field and I'm already developing some other materials along this line that I could double-dip :-D

Other possibilities:

• There are lots of potential public-health type datasets. I was recently engaging with police violence datasets compiled by various organizations, though this may be considered too political and/or triggering. There are also lead exposure / air quality datasets I'm aware of
• Climate / weather data is very relatable and evergreen.
• Lots of potential ideas from https://www.wikidata.org