COMM 207

Assignment 3
Module 8
Due: Sunday December 1st 11:59 PM (submit via Canvas)

Instructions (please read carefully!)

Please submit your assignment via Canvas by the deadline. Instructions are given below. If the
marker cannot read your work, then you will receive a grade of 0. If a question says to use Excel,
then do not submit handwritten work for that question. Yes, you may use your tablet and stylus if
you wish. Your assignment must include a signed declaration of academic integrity (provided
below – type or write your name into both fields). Failure to submit a declaration of academic
integrity with your name in both fields will result in a grade of 0% on the assignment. A
video showing how to upload your assignment has been posted to Canvas > COMM 207 >
Modules > Start Here.

Please DO NOT SUBMIT YOUR EXCEL FILE. Submit only ONE document – a PDF with
your answers and your declaration of academic integrity. Rounding instructions are given in
each question.

How to export your assignment from Word to PDF if you are using a Mac

File > Print… > PDF (bottom left drop-down menu) > Save as PDF…

How to export your assignment from Word to PDF if you are using a Windows computer

File > Export > Create PDF/XPS (check the pop-up window says “Save as: PDF”)

How to upload pictures of your handwritten work to Word

a. Save each picture as a .png or .jpg file.

b. In Word, go to Insert > Pictures > Picture from file, and select your first picture.
c. Ensure the picture is large enough to be read on a screen, and the image is not blurry.
d. Proceed to the next picture, starting from step a., until all pictures are contained in the Word
document. Save the document, convert it to PDF, and submit the PDF file on Canvas.

What to do if you encounter technical difficulties

Update Google Chrome - click the three vertical dots in the top right corner, and then click
Settings > About Chrome. Chrome will automatically update if there are any updates
available. Then, clear the browser history and cache. You can follow the instructions from the
link here: How do I clear my web browser cache/history? Then, log out of PAWS and close
Chrome entirely (don't just close the window!) and re-open it.

If you still encounter technical difficulties after following the above steps, please email me at
least 48 hours in advance of the deadline for assistance.
Declaration of academic integrity (MUST be included in your assignment submission)

I, Dawsar Abdelhadi (typed name as it appears in Canvas) declare that I did not use any
unauthorized resources when completing this assignment.

I understand that the authorized resources for this assignment are:

- The posted COMM 207 material on Canvas

- The tutorials
- The YouTube videos linked to from the course syllabus posted on Canvas
- Office hours
- The Math & Stats Learning Centre
- Discussions with a classmate: I understand I may tell a classmate where to find an answer to a
similar question in the course material, but I must not share my answer.

I understand that posting these questions, or my answers, or the instructor’s answers to a website
or app will result in academic misconduct charges.

I understand that use of an unauthorized resource will result in academic misconduct charges.

I understand that academic misconduct charges carry a penalty of anywhere from 0% on the
assignment to expulsion from the university.

Signature (type name again here): _Dawsar

Abdelhadi_______________________________________
Assignment Questions

1. [4 marks] Explain when each of the two chi-square tests covered in Module 8 is used, and
provide a quick reminder to yourself about some key qualities about the chi-square distribution.
Fill in the blanks below to answer the question.

I should use the chi-square test of goodness-of-fit when I want to know _if value is different
from its hypothesized value________________

_____________________________________________

I should use the chi-square test of independence when I want to know _if 2 variables depend on
each other ________________

_____________________________________________

I can only use one of these two chi-square tests on this type of data (one word):
_categorical__________

The chi-square distribution is (pick the correct shape from this list: left skewed, right skewed,
symmetric): _ right skewed_______________

Case A [12 marks]: The Bank of Canada increased interest rates a few times in 2023 [1].
Suppose a survey was conducted by a polling organization, and Saskatchewanians answered the
question: “have you been majorly affected, somewhat affected, or not at all affected by interest
rates climbing?”. The survey yielded the following results.

Majorly affected Somewhat affected Not at all affected

700 280 50

The same survey was conducted in Ontario, and it was found that 69% of Ontarians believed
they would be majorly affected, 29% believed they would be somewhat affected, and 2%
believed they would not be affected at all. The polling organization is interested in determining if
there is a difference in the population proportions between the two provinces. Conduct an
appropriate hypothesis test by filling in the blanks below. Let group 1 be the “majorly affected”
group; group 2 be the “somewhat affected” group, and group 3 be the “not at all affected” group.

CHI TEST OF MULTINOMIAL

1) [1 mark] State the null and alternative hypotheses using the appropriate notation.

HO: P1=0.69, P2= 0.29, P3=0.02

HA: at least one 𝑝𝑖 is different from its hypothesized value, for i =1, 2, 3
2) [1 mark] Calculate the test statistic. Please round your final answer to 2 decimal
places. Note: you may use Excel to calculate the test statistic. No need to show your
work. Hint: Note it’s not necessary for expected values to be whole numbers, as
these are essentially averages.

N= 1030

1030(0.69) = 0.161094695
+

1030(0.29) = 1.170706394
+

1030(0.02) = 41.9592233

43.30

3) [1 mark] Find the p-value using Excel. Use your rounded test statistic from part 2, and
show your Excel command. Please do not round your answer. For example, suppose this
question had instead asked you to calculate a sample mean of 10 values. If I had data in
cells A1 through A10 in Excel, and I typed =average(A1:A10), hit enter, and got the
result 10.2, then I would provide: =average(A1:A10)=10.2.

=CHISQ.DIST.RT(43.30,2)

= 3.95844E-10

4) [1 mark] Based on this p-value, do you reject the null hypothesis (answer “yes” or “no”
only, with no additional words)?
yes

5) [1 mark] In one sentence, conclude in the context of the original question.

There is no statistically significant difference in the population proportions of categories of
interest rate climbing in the two provinces

6) [7 marks] Conduct a follow-up analysis. Show all of your work. Do exactly as shown in
the Module 8 Tutorial, including all rounding (copy/paste from that document and change
numbers and context). Ensure you have the entire follow-up analysis (i.e., everything on
that page, not just the table). Note: you’re going to get a strange answer here (you will
end up rejecting only one of the null hypotheses, which doesn’t agree nicely with
intuition). This was discussed at the end of the Module 8 Review Questions and answers.
It’s not a mistake.

1. Find each sample proportion ˆpi

P1: 700/1030= 0.67961165

P2: 280/1030= 0.27184466

P3: 50/1030= 0.048543689

2. State the sampling distribution of the sample proportion

QP1=

Square root of: 0.67961165 (1−0.67961165) / 1030

≈ 0.01454

QP2=

Square root of: 0.27184466 (1−0.27184466) / 1030

≈ 0.01386

QP3=

Square root of: 0.048543689 (1−0.048543689) / 1030

≈ 0.00660

 So the sampling distribution of each sample proportion is:

pˆ1 ∼ N(0.67961165, 0.014539523)

pˆ2 ∼ N(0.27184466, 0.013862891)

pˆ3 ∼ N(0.048543689, 0.006696413)

3. Calculate the confidence interval using formula

 pˆ1 ± zα/2 r pˆ1(1 − pˆ1) n = 0.67961165 ± 1.96(0.014539523) ≈ [0.65111, 0.70811]

 pˆ2 ± zα/2 r pˆ2(1 − pˆ2) n = 0.27184466 ± 1.96(0.013862891) ≈ [0.244673394, 0.29902]

 pˆ3 ± zα/2 r pˆ3(1 − pˆ3) n = 0.048543689 ± 1.96(0.006696413) ≈ [0.03541872, 0.06169]

Case B [9 marks]

a) [5 marks] A university is interested in examining the relationship between the secondary

school at which a student studied and their first choice of college within the university. The point
of the study is to target program marketing to students who may not be aware of certain
programs. They conduct a study and collect the following data. Conduct an appropriate
hypothesis test using the critical value method. Round your test statistic to 2 decimal places. Do
not show your work regarding how you find your test statistic.

Secondary school 1 Secondary school 2 Secondary school 3

College 1 1500 500 500
College 2 1500 500 500
College 3 1100 350 300

chi-square test of independence

1. H0: Preferred college is independent of secondary school.

HA: Preferred college is not independent of secondary school

2. Test statistic:
 7.23667

3. critical value:

=CHISQ.INV.RT(0.05,(3-1)*(3-1)) = 0.99969266

4. Since the test statistic (7.23667) doesn’t lie in the rejection region (9.48773, ∞), we don’t
reject the null hypothesis.
5. The variables “college” and “secondary school” are independent.

b) [2 marks] Based on your analysis in part a), do you think we can use the variable “secondary
school attended” to predict which college a student will choose as their first choice (answer
“yes” or “no” – one of these words must be present or you will lose a mark)? Why or why not?

No we cannot use the variable “secondary school attended” to predict which college a student
will choose as their first choice.
c) [2 marks] Think about the real world for a moment. Give one reason why your answer to part
might make sense, and one reason why it might not. Answer in 2 sentences.

Human beings are all different and make choices based on external factors. For example,
personal interests and life situations all differ for every person.

References

[1] Bank of Canada. Accessed 17 December 2023. Available:

https://www.bankofcanada.ca/core-functions/monetary-policy/key-interest-rate/