Assignment #3 has been graded and e-mailed back to the person who sent it to me. Please examine the comments that I made in red using Word’s tracking tool or with my own writing (turn on “all markup” to make sure you can see these comments). If you have any questions about how to do a particular item in the assignment, please feel free to ask during class, post a message to the listserv, or stop by my office to ask. If you have a question on why I took off a particular number of points for a part of your assignment, please ask me individually after class or stop by my office. 

I have posted an answer key to the assignment on the “Graded Materials” web page of the course website. Please look over all parts of it and compare my answers to your answers – even if you got a problem correct. 

There were a total of 44 points possible. The mean was 38.4 points, the standard deviation was 1.03, and the high score was 39.5. Below are a few comments:

1b: It appears Rauser was maybe referring to our class too :(. Please see my answer. 

1ci: I was kind of surprised to see no one use the t.test() function within the calc.t() function, especially because everyone used the t.test() function for 1a. Of course, you can get the correct answer without t.test(), but one might as well use it to make everything easier. 

1ci: This part and some other parts resulted in a p-value of 0.0004. This value can only occur when the test statistic is the most extreme. If a larger number of resamples were taken, this may not happen. Therefore, one will usually say the “p-value is < 0.0004” because we are limited to only R = 4999 resamples here. 

1c iv and 1di: A few students decided to take resamples from within the calc.t2() function like in a double bootstrap. This is not necessary here. The boot() function takes the resamples you need to perform the test. 

After 1e: Please see the last plot that I made in the assignment. It gives an interesting look at what the models under the null hypothesis look like.