- Sam Penrose | http://www.sampenrose.net | twitter.com/sampenrose
I. Welcome! (5 minutes)
- We're going to learn: - what "data" is - what "data analysis" is - the key ingredient whose absence ruins otherwise good work
- Open a new browser window and load https://git.io/datatalk (this page).
- In a second tab, create an empty Google spreadsheet - You'll need a Google account. - Alternately, you may work in a different spreadsheet application.
II. What is data analysis? (5 minutes)
- A warm-up exercise.
- Data analysis is organized thinking.
- Data analysis is a spiral: + categorize -> count -> compare -> communicate -> consider -> (and perhaps loop back to an earlier step)
- Data analysis is a skill you already have.
III. Count and compare: analysis of a real dataset (15 minutes)
- In a new tab, open Lena Groeger's wonderful Spreadsheets Lab . We'll do it together. * In Step 5, add a second filter: GRADE DATE is after 09/01/2015 . * In Step 16, summarize CAMIS by COUNT (not COUNTUNIQUE) and choose a pie chart (not bar chart).
IV. Communicate: what did we learn? (5 minutes)
- Your responses.
- Do we need to loop back? If so, to which step?
V. Consider and (back to) categorize (15 minutes)
- Maybe a second dataset will shed more light on our concern. * Create a new spreadsheet of food poisoning data * Form pairs to answer: "Can we find evidence of a problem with our choice?" (10 minutes)
- Is our choice in this dataset?
- What makes a restaurant dangerous? * Incidents -> reported? * Grades -> not "A"? * Issues -> "critical"?
VI. Discussing what we learned (5 minutes)
- What is the key ingredient needed to drive good data analysis? * Hint: see IV.2 and V.3
- Can we ever skip the key ingredient?