Week 2: Describing Data

[andrewpbray]

How it went

Themes

• Describing Data is often a process of summarization - what to include and what to omit.
• Description can take verbal, numerical, or graphical forms.
  • Verbal: Describing distributions: modality, skew, what's a typical observation? , how spread out is it?, what's the maximum?
  • Summary statistics: designed for formalize the way people answer the description questions
  • Graphics: information loss, perception /

Practicalities.

• Probably need two "Describing Data" (potentially one tied to Ch. 4, one to Ch. 5) lectures, one "Grammar of Graphics", one "Exploratory Data Analysis / Data Viz in practice".
  • Describing Cat data: counts, proportions, barplots
  • Describing Numerical data: quartet of graphics, summary statistics statistics
• "Grammar of Graphics" lecture needs to come the day before a lab where they practice.
• For the "Grammar of Graphics" lab, do use data collected from students, but you'll need to go through IRB to host it publicly.
  • What major do you intend on pursuing?
  • What year are you?
  • Have you programmed before?
  • How are you feeling about stats / programming? Likert Scale
  • Pants on dog