Create mini project/analysis 2

[wvictor14]

This is for milestone 2 of the mini analysis project.

The topics covered are

• dplyr::group_by()
• dplyr::summarize()
• advanced ggplot2
• a brief section of tidyr reshaping

I tried to make the dplyr and ggplot2 sections more difficult than milestone 1. Hope it is not too difficult! I allocated 20 points to this section.

The tidyr section is brief, with the understanding that they might get more practice in milestone 3. But I'm also not sure exactly what milestone 3 is about.

I'm also not sure if the tidyr section is too open-ended. I looked at the datasets from the datateachr package, and it seems that they come in a variety of formats, some more tidy than others. It was hard for me to think of specific questions. Let me know what you guys think!

Victor

[wvictor14]

Can also reallocate points such that 5 / 30 are reserved for reproducibility, readability, etc. Not sure where we landed on that from the PR for milestone 1

[iciarfernandez]

Added a few minor comments - looks good to me! @wvictor14

Re: Difficulty - I think they should definitely be able to complete the tasks that you've listed by the time this milestone is due.

Re: Tidy data & Milestone 3 - I don't think the tidyr section is open-ended, but I would maybe add one last point for them to narrow down their research questions to, say, 2 out of those 4 (maybe even 1?), and then subset their data to a tidy dataset that they will use in Milestone 3 (with the idea that Milestone 3 will cover more of ggplot2, and perhaps modeling at a high level) - therefore, they can choose to drop variables that they are not interested in investigating at this stage. This can also double as a mini-tutorial to saving intermediate objects - I'm not sure if it will be covered in the course, but I think it's simple enough to ask them to save this table as either an RDS object or other format in their repos, and then teach them how to read it in for the next Milestone.

Re: Points - I don't think we have agreed on anything yet? Discussion is happening on #56. Maybe @vincenzocoia can chime in on this one, but yeah I think it would be nice to allocate a few points to reproducibility, coherence, etc.

[wvictor14]

don't think the tidyr section is open-ended, but I would maybe add one last point for them to narrow down their research questions to, say, 2 out of those 4 (maybe even 1?), and then subset their data to a tidy dataset that they will use in Milestone 3 (with the idea that Milestone 3 will cover more of ggplot2, and perhaps modeling at a high level) - therefore, they can choose to drop variables that they are not interested in investigating at this stage. This can also double as a mini-tutorial to saving intermediate objects - I'm not sure if it will be covered in the course, but I think it's simple enough to ask them to save this table as either an RDS object or other format in their repos, and then teach them how to read it in for the next Milestone.

OK sure, yeah that makes sense! And thanks for the feedback 👍

[vincenzocoia]

Re: Points - I don't think we have agreed on anything yet? Discussion is happening on #56. Maybe @vincenzocoia can chime in on this one, but yeah I think it would be nice to allocate a few points to reproducibility, coherence, etc.

Yeah, some points for this sounds reasonable to me.

[wvictor14]

Grading scheme:

• [ ] 2.5 for reproducibility, readability, and repo organization (RRR)

• [ ] 15 for Tasks 1.1-1.3

  2.5 for listing 4 research questions from milestone 1 10 for summarizing and graphing 2.5 for reflecting on their findings

• [ ] 12.5 for Tasks 2.1 - 2.3

  2.5 for identifying if their data is tidy or untidy 5 for tidying their untidy data (pivot_longer), and then back to their original state (pivot_wider) 5 for picking 2 research questions, and explaining their reasoning based on their findings in this milestone. Also for saving a version (altered in some way we've covered) as an .rds file

[dy-lin]

hey @wvictor14,

@sciclic and I created this 'one-glance' checklist for the students to look over before they do a submission. This was implemented after student feedback about having to look in way too many places for critieria. (e.g. TB1 had criteria in the document that was not in the README, etc). See the example we sent out for milestone1:

The MD file: https://github.com/UBC-STAT/stat545.stat.ubc.ca/blob/master/content/mini-project/mini-project-1-checklist.md

MD file as body of an issue: https://github.com/stat545ubc-2020/template-mini-analysis/issues/2

I created an issue on each of the repositories with this checklist! Things I didn't add but think would be good:

• the stuff about repo organization, updating README's, knitting, submission, etc., especially the stuff that isn't for marks (I was focusing on things for marks on this last checklist)

For milestone1 we created the issue today as not all the repositories existed yet, but for milestone2 we were thinking we could create the issue at the same time as we're creating the pull requests with milestone2 documents to their repositories.

I guess you'd basically be converting your previous comment above to a possibly more detailed breakdown .md file with a checkbox for each point so they can check it off in their issue. Let me know if you'd like me to do this (convert your comment above) if you have other stuff to do

[vincenzocoia]

I'm going to merge this PR now. I think everything looks great, though if we wanted to add a checklist (would be a good idea in response to student feedback), we can do that sometime this week.

Thanks for the activity on this pull request!

[dy-lin]

@wvictor14 I had some time, so I combed through the .Rmd and created this checklist: https://github.com/UBC-STAT/stat545.stat.ubc.ca/blob/master/content/mini-project/mini-project-2-checklist.md

While making it, I noticed that the point system is different from the what you commented above, where we added points for RRR:

in the Rmd:

Task 1 (17.5 pts)

• 1.1 (2.5 points)
• 1.2 (10 points)
• 1.3 (5 points)

Task 2 (15 points)

• 2.1 (5 points)
• 2.2 (5 points)
• 2.3 (5 points)

above comment:

Task 1 (15 points)

• 1.1 (2.5 points)
• 1.2 (10 points)
• 1.3 (2.5 points)

Task 2 (12.5 points)

• 2.1 (2.5 points)
• 2.2 (5 points)
• 2.3 (5 points)

RRR (2.5 points)

@vincenzocoia I see there is new spot in the table on Week 5 Guidance for the checklist! I'll link it there once we sort out the point system?

[wvictor14]

@dy-lin Oh shoot I did not push my last commit, I'll do that now, Thanks for spotting that