1. We were interested in finding out whether a drug could affect the weight of rats (measured in grams). We first drew a simple random sample of 58 rats, and then randomly assigned each rat to either the treatment group or the control group. After the assignment, there were 29 rats in the treatment group and another 29 in the control group. We administered the drug to all rats in the treatment group, and administered the placebo (i.e., substance without any effect) to all rats in the control group. Then we raised the two groups of rats in the same environment for a month. After a month, we weighed the rats in both groups. This investigation is best described as a

(a) cluster sample

(b) experimental study

(c) quota sample

(d) observational study

(e) None of the above

2. Which of the following is NOT an assumption in multiple linear regression?

(a) Independence of samples

(b) Normality of errors

(c) Equal variance of errors

(d) n > 30

(e) None of the above

3. We are interested in determining if the average fuel efficiencies of U.S. cars with manual and

automatic transmissions differ. Data on average highway mileage (measured in miles per gallon) of 15

randomly selected car models each with both a with manual and automatic transmission version. For each

car type, miles per gallon were recorded for manual and automatic transmissions of the 15 matched (make,

model, model year, and body style) versions. In order to carry out the comparison you need a test of:

(a) Unpaired means

(b) Two sample proportion

(c) Paired means

(d) One sample mean

(e) Bonferroni correction