Communicating data science results effectively

Lecture 24

Dr. Mine Çetinkaya-Rundel

Duke University
STA 199 - Fall 2022

11/29/22

Warm up

Announcements

• Office hours today 2:45 - 4:45pm in person or on Zoom

• Remaining due dates:

  • Peer eval 2 due tonight by 11pm. Late submission is not possible!

  • Survey on Sakai for topics requested for next week.

  • Project presentations next Monday in lab – all team members must be present.

  • Project write-up and final repo due 11:59pm Thur, Dec 8 – you will lose access to your repo at this time.

  • HW 6 due 11:59pm Friday, Dec 9.

  • Remaining application exercises due on the usual schedule.

  • (Optional) Exam retake – due 11:59 pm Thur, Dec 15 – absolutely no late submissions, extensions, etc. for any reason.

Exam retake

• Released after Exam 2 grades are released, by next week.

• Weighted average between your lowest exam score and what you earn on the retake

• If you earned a 0% on Exam 2 and earn 100% on Exam Retake, you will earn an 85%.

• Overall task: Create a six question assessment covering specific topics from the course, write a key and a grading rubric for each, as well as a justification.

Project

• Review peer evaluations left by your peers, implement updates as you see fit, close the issue once you review them.

• Have a clear plan for who is doing what, open issues on your repo, and assign them to individuals who can then close the issues as they finish a task.

• Schedule at least one team meting between today and your presentation to practice your presentation together.

Effective communication

What’s going on in this plot?

Take A Sad Plot & Make It Better

Source: https://alison.netlify.app/rlm-sad-plot-better

Application exercise

ae-20

• Go to the course GitHub org and find your ae-20 (repo name will be suffixed with your GitHub name).
• Clone the repo in your container, open the Quarto document in the repo, and follow along and complete the exercises.
• You should have already pushed updates on Tuesday, so you should be good for submission.

Recap

• Represent percentages as parts of a whole
• Place variables representing time on the x-axis when possible
• Pay attention to data types, e.g., represent time as time on a continuous scale, not years as levels of a categorical variable
• Prefer direct labeling over legends
• Use accessible colors
• Use color to draw attention
• Pick a purpose and label, color, annotate for that purpose
• Communicate your main message directly in the plot labels
• Simplify before you call it done (a.k.a. “Before you leave the house, look in the mirror and take one thing off”)