assignments

1. Assignment 1 due Fri Feb 16

From OpenIntro Stats (p55 onwards): 1.4, 1.11, 1.28, 1.61.
Choose a cookie from the cookie data. Calculate the standard deviation of its scores by hand.
Do a better job than the rent/wage meme on the prep page for this week.
Play with R and ggplot and let me know how you're doing. Include at least one ggplot-made graph.

From OpenIntro Stats (p158 onwards): 3.4, 3.6, 3.14.
From OpenIntro Stats (p203 onwards): 4.8, 4.10, 4.18.
Do something cool in R and/or ggplot2.
Do one of:
- Find a somewhat complex graph that you think does a (mostly) good job. Discuss what it is telling us, why you think it's doing it well, and anything you think it could do even better.
- Select some exercises from Chapter 2 of OpenIntro Stats that demonstrate what you've learned from it. Answer them.

From OpenIntro Stats (p203 onwards): 4.24, 4.26, 4.28, 4.40.
From OpenIntro Stats (p257 onwards): 5.8, 5.18, 5.28.
From the probability notes on the closed resources page, do some exercises from the first three sections. If you're already comfortable with probability, do harder ones; if not, do easier ones.
Bonus question: 3.3 from the probability notes: design and analyse a pig-like game.

From OpenIntro Stats: 5.43 (p. 269).
Look at the "cars" data set in the openintro package. Use an ANOVA test to determine whether different types of car (small/midsize/large) have different mileages (stored in the mpgCity variable of the data set). Write up your answer in a way that conveys the important points of your analysis and conclusions to a non-statistician.
Find one or more datasets out in the wild (it does not have to be very wild; there are lots of R packages that consist solely of data that is in ready-to-read-and-analyse format). Perform t-tests and/or ANOVAs on appropriate ones. (You'll be getting this question again with different tests on future assignemnts, so keep note of cool data sets even if they don't fit the criteria for t-testing or ANOVA.)

From OpenIntro Stats: 6.45 (p.324), 7.11 (p.358).
Find one or more datasets out in the wild (it does not have to be very wild; there are lots of R packages that consist solely of data that is in ready-to-read-and-analyse format). Perform \(\chi^2\) tests and/or linear regressions on appropriate ones.

I'll share the exam with everyone via email at 12noon on Thursday 3rd May. Due 24 hours later at 12noon on Friday. I'll be around for questions in person for the first couple of hours and the last couple of hours (and probably more) and also as available as possible via email over the time the exam is active.