The course website contains the sids.txt data set that lists the sex and ages at death (in days) of 16 children who died of sudden infant death syndrome in Seattle from 1973-1982. We analyzed the weights of these kids. This time, we are interested in their age at death.

"Sex" "DeathAge" "F" 60 "M" 78 "M" 80 "M" 77 "F" 87 "M" 115 "M" 175 "F" 56 "F" 60 "M" 114 "M" 81 "M" 58 "M" 103 "M" 134 "M" 46 "F" 53

(a) developing a table that lists the mean and standard deviation for age at death for males and females, along with the sample size for each group. Based on the table, would you expect Student's t-test and Welch's t-test to give similar answers?

(b) Test whether or not the difference in age at death for males and females could be due to chance using Student's t-test.

(c) Test whether or not the difference in age at death for males and females could be due to chance using Welch's t-test.

(d) Construct a 95% confidence interval for the difference in age at death for males and females, based on the assumption of equal variability in males and females.

(e) Construct a 95% confidence interval for the difference in age at death for males and females, without assuming that variability is equal in males and females.

(f) Which approach (Student/Welch, equal/unequal SD) do you think is most appropriate in this circumstance? Why?