Homework 1 DS502/MA543

Homework 1
DS502/MA543
So that everyone in the class has the maximum access to background material Homework 1 will focus
on problems from the book. Your answers will be graded both on your demonstrated understanding
of the concepts from the book and from the class, as well as the clarity of your explanations.
For example, given a question such as:
Describe the null hypothesis to which the p-values given in Table 3.4 correspond. Explain what
conclusions you can draw based on these p-values. Your explanation should be phrased in terms of
sales, TV, radio, and newspaper, rather than in terms of the coefficients of the model.
An unacceptable, though perhaps strictly true, answer would be:
It appears the TV and Radio matter more than newspaper.
An acceptable answer would be:
The small p-values in Table 3.4 for TV and radio indicate that, holding the other predictors constant,
each of these predictors will, with high likelihood, change the amount of sales. For example, fixing the
amount of TV and newspaper advertising, while increasing the amount of radio advertising, will very
likely lead to an increase in sales because the p-value in Table 3.4 is very small for radio and radio's
coefficient is large and positive. On the other hand, since the p-value for newspaper is quite large, the
data indicates that newspaper advertising is unlikely to have any effect on sales when TV and radio are
held fixed. Note, Table 3.3 does show a small p-value for a single linear regression across newspaper
and sales, but this is likely because newspaper advertising is predictive of TV and radio advertising,
not necessarily because newspaper advertising directly influences sales.
Please submit your homework online on canvas.wpi.edu before the start of class on Tuesday,
February 6. Please make sure that both team member names appear on the submission and that every
submission is stand-alone (i.e., does make the grader read or run your code) and don’t zip your
files! In addition, for any problem on which you write code, plus submit your code online on
canvas.wpi.edu before the start of class on Tuesday, February 6. You will not be graded your
programming style, but having access to your code will allow us to more easily give partial credit.
You are also welcome to use any of the R scripts in the book. To keep things organized, please send
your R files to Haitao and Brendan using the following naming convention:
<Last name person 1>_<First name person 1>_<Last name person 2>_<First name person 2>_HW1_<question number>.R
For example, if Haitao and Fatemeh were a team, the file containing the code for problem 4 would be
named:
Liu_Haitao_Emdad_Fatemeh_HW1_4.R
Homework questions
1. (10 points) Section 2.4, page 52, question 1
2. (10 points) Section 2.4, page 52-53, question 3
3. (5 points) Section 2.4, page 53, question 6
4. (15 points) Section 2.4, page 54-55, question 8
5. (10 points) Section 2.4, page 56, question 9
6. (5 points) Section 3.7, page 120, question 1
7. (10 points) Section 3.7, page 121, question 5
8. (5 points) Section 3.7, page 121, question 6
9. (15 points) Section 3.7, page 121-122, question 8
10. (15 points) Section 3.7, page 122, question 9 (Note, this question is the most open ended of the
R programming questions, in particular parts (e) and (f). The book will provide insights into
what needs to be done here. See how far you can get.)