BUS232 Spring 2024 A4

Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1of 4

BUS232, Spring 2024

Assignment #4
Instructor: Negar Ganjouhaghighi

Question Mark Maximum


1 100
Total 100

Due: March 24th , 2024 by 11:59 pm on Canvas


Submit one Pdf file which includes the answer for all questions and one excel file

Group Number:
(mandatory)
(group #)

Last Name First Name


(in alphabetical order)

The names and group number information in the grid above must be filled in correctly or
marks will be deducted. Group information is posted online.
BUS232 (Spring 2024)
Business Statistics

ASSIGNMENT #4
Central Limit Theorem, Confidence Interval

1. Question 1 [100 marks]

For this question, we want to recreate the following graph.

To do so, we need to have 3 populations with a distribution other than normal distribution and a
population with normal distribution.
Attached you can find 4 populations with Normal, exponential, Poisson, and uniform distributions.
Normal and Uniform populations are just simulated data while the exponential distribution is actual

BUS 232 (Spring 2024) Ass’t #4 Page 2


data for the time between two patient assessment by a physician and the Poisson distribution is the
hourly number of arrivals to an ED in Calgary (From my thesis).
(Here is what we want to do: for each population, we want to find the average of 200 samples. First,
we take 200 samles, each size of 2 (only 2 numbers in the sample). Then another 200 samples with
size of 5. And lastly 200 samples of size 30. Then we find the sample average for each sample: for 200
samples of size 2: we will have 200 averages. For the 200 samples of size 5: we will have 200 averages
and the same for 200 samples of size 30: 200 averages.
Then we insert one histogram for each sample size: for example: for sample size 2: we have 200
numbers, we select them and then insert a histogram )
To graphically prove the central limit theorem, you need to use the random sampling technique and
take multiple (say 200) samples of sizes 2,5, and 30 of each of the populations.
For each population and sample size n:
a. Create 200 samples of size n
b. Calculate the mean for each of the samples. You will have 200 sample means.
c. Then draw the histogram of these 200 numbers.
So, you need to do the above steps 4 (population) x 3 (sample sizes of 2,5,30) or 12 times.

In order to create random samples in excel:


I. Copy and paste the following formula to the blue cells of each sheet:
a. =INDEX($A$2:$A$251, RANDBETWEEN(1, ROWS($A$2:$A$251)), 1)
b. The first input is the range containing your population data. For me it is column A rows
2 to 251. Make necessary changes so that you are including all the population.
c. The second input is a random number from 1 to the last row of your population data
point. You need to adjust this as well.
d. The index (Range, a,b) function output is the value you have in the ath row of the
range. The last input, b, is he column number, which is 1 here as we only have one
column.
e. Once you are done with randomly selecting your samples, copy them all and paste as
values so that they don’t change every time you change something in excel.
II. The green cells are the average of the blue column above them. Or, they are the sample
means for each of the samples.
III. In each sheet, you have 3 green rows, containing the average of your samples with sizes 2, 5,
and 30. Draw one histogram for each of them. You can change the number of bins as required.

95% confidence interval: (Orange cells)


We want to construct a 95% confidence interval for sample of size 30 for each population.
- For each sample and each population: total of 200(samples of size 30)*4(population)=800
samples
- In the rows labeled Lower Bound and Upper Bound: find the LB and UB for each sample.
- In the row below them: write the actual population average. (will be the same for all samples)
- In the row below that: write “ERROR” if actual population average is not in the confidence
interval created above and “OK” otherwise)
- Find the probability of “ERROR” happening by dividing the number of “ERROR”s to 200.

BUS 232 (Spring 2024) Ass’t #4 Page 3


The output of this question:
1. Submit your excel file that contains your samples, histograms and confidence intervals. You
should have a sheet for each population. In the same sheet, there should be all of the samples,
one column for each sample size.
2. In the word (which is then converted to a pdf file) file create a 4x4 table and copy and paste
the histograms from excel in the corresponding cell of the table. Exactly the same as the table
above. There should be total of 16 histograms, which means that you should include the
histogram of populations as well.
3. Fill out the table below:

Average of Average of Average of


Standard 200 sample 200 sample 200 sample
Population Average P(Error)
Deviation averages averages averages
with size 2 with size 5 with size 30
Normal
Exponential
Poisson
Uniform

BUS 232 (Spring 2024) Ass’t #4 Page 4

You might also like