🔔Lecture 5 - Sampling distributions

1. A statistic from a sample has its own distribution: the one you would see if you repeated the sampling many times.
2. The 🔤 Standard error is the spread of that distribution. Standard deviation describes packets. Standard error describes means of trays.
3. It shrinks with the square root of the sample: four times the packets, half the wobble.
4. Second attempt, 1 June: the simulation in problem set 2 is what did it. I had to watch the pile form.

Goes with

• 📌 Statistics I - Midterm
• 🧪 Spaced review of lecture notes
• A-B test sample size sheet
• Statistics flashcard deck

• today's idea: the sample mean is itself a random thing. take another tray, get another mean
• imagine ALL the trays you could have taken. their means make a pile
• the pile is narrower than the packets. much narrower
• width of the pile = 🔤 Standard error. NOT standard deviation (I wrote SD four times and crossed it out four times)
• SE = s / √n
• 20:25 central limit something. the pile is bell-shaped even when the packets are not?? lost it here
• 20:50 nodding. not following