ASSIGNMENT - 1

Nagarajan Chinnakani (2318620)

University Canada West

BUSI 650: Business Analytics (HBD-WINTER 24-62)

Dr. Behnaz Bahmei

Feb-03-2024
A. Hypothesis Testing

Background: A fast food restaurant claims that the average wait time for a drive-thru order is 3
minutes. The restaurant manager wants to test this claim with a sample of customers.
Task: You have been hired as a data analyst to help the restaurant manager with hypothesis
testing and confidence intervals.

1. Formulate the null and alternative hypotheses for this situation.

Null hypothesis: The Null hypothesis ( H 0), is that the fast-food restaurant claims,

i.e. the average time for a drive-thru order is 3 minutes.

H 0 : μ=3

Where µ is the population mean wait time for a drive-thru order.

Alternative hypothesis: The Alternative hypothesis ( H a ), is the opposite of what

the Null hypothesis states, i.e. the average time for a drive-thru order is not equal

to 3 minutes.

H a : μ ≠3

2. If a random sample of 50 customers produces a sample mean wait time of 2.8 minutes

with a sample standard deviation of 0.4 minutes, calculate the test statistic using a z-

value.

x−μx
To calculate Z – Statistics, Z = σ
√n
 Where, σ is the sample standard deviation, 0.4 minutes.

n is the sample size, 50 customers

x is the mean wait time, 2.8 minutes

μ is the population mean wait time, 3 minutes.

2.8−3
Therefore, Z – Statistics, Z = 0.4
√50
Z – Statistics = -3.536

3. Calculate the p-value for this test and interpret the results at a significance level of 0.05.

Using the Z – Value -3.536, the obtained P – Value is 0.000203 (from online

calculator)

Since the population mean wait time is not equal to 3 minutes, this test should be

performed by two tailed test, Hence doubling the P – Value, we get 0.000406

By Interpreting the P – Value to the given significant value α

i.e. P -Value, 0.000406 ¿ 0.05 ∝

hence, we reject the Null hypothesis, since the P – Value is less than the

significant value α . So there is evidence that the average wait time is not 3

minutes as claimed by the Fast food restaurant.

4. Determine the critical value for a one-tailed test with a significance level of 0.05 and

compare it with the calculated z-value. What is your conclusion?

By using the Critical value calculator,

The critical value for a one-tailed test with a significance level (α ) 0.05 is -1.6449

We have the calculated Z – Value of -3.536

While comparing the Z - Value of -3.536 to the critical value of -1.6449, the

calculated

Z – Value is less than the critical value.

Hence, we reject the Null hypothesis on the one-tailed test with a significance

level of 0.05. It states that the average wait time for the drive-thru is less than 3

minutes, and it proves that there is no evidence to support the claim of the fast

food restaurant.

5. How would decreasing the sample size from 50 to 25 affect your conclusion? Explain.

Decreasing the number of sample sizes from 50 to 25 would make it much

difficult and challenging to obtain a clear conclusion, because less sample sizes

might have more flexibility in their mean values which mostly fails to present the

overall population mean value. For example, if the smaller sample sizes consist

mostly of the minority greater values and fails to include the majority of other

values, would bring higher differences in the mean value, hence having a larger

sample size would helps to analyzing to give a better confident result.

6. Discuss potential sources of error or bias in the fast-food restaurant example.

The following are some of the possible errors which might affect the sample size.

 The recorded sample data might not be recorded at the perfect time.

 The presented sample data might not cover the entire sample.

 The data observed time might affected due to weather conditions or traffic at the

zone.

 There might be a shortage of manpower at the serving point during the data

observed.

 The sample collected may be on a busy day.

The possible errors can be mitigated by well analysing the restaurants pattern before

observing the data and also by gathering a large sample size.

7. Provide two other examples of situations or industries where hypothesis testing and

confidence intervals could be useful, and briefly explain why.

Sports Sector:
In the field of sports Sector, the coaches, team administrators and sports
analysts use hypothesis testing to assess the performance of their players and to
measure the effectiveness of various training techniques employed, or to compare
strategies used by different teams.
When it comes to performance measures, confidence intervals may be
applied to demonstrate their sportsmen’s reliability and consistency in playing
their games.
 Educational Sector:
Hypothesis testing in educational sector is a key component of evaluating
the effectiveness of a teaching strategy or analysing the instructional approach in
the field of education.
By applying the confidence interval, the institute or the educational sector
may update or improve their techniques and strategies in teaching students to get
a higher result.

B. Moving Average

Imagine you are a consultant hired by a retail store looking to enhance customer satisfaction. The
management wants to employ data-driven strategies to understand customer preferences,
improve service, and ultimately boost sales. Your task is to design a comprehensive data
collection and analysis plan.
a) Identify both numerical and non-numerical data relevant to understanding customer
satisfaction. Explain how each data type contributes to the overall analysis.
Numerical Data:
 Sales Figures: the numerical data which gives the sales value per day. This data
helps to find the sales pattern and the customer behaviour.
 Customer counts: this is numerical data which helps to find the number of
customers reaching the business daily and can be used to conduct a survey of
customer ratings.
Non-Numerical Data:
 Product types: this is a data about the products which the customers mostly
prefer to buy or do not like to buy, which helps the business to increase or
improve the specific product or service.
 Customer Reviews: this data is the satisfaction level from the customers,
mentioning whether the business is providing services good or bad.

b) Propose a plan for random sampling to ensure a representative subset of customers is

included. Suggest how stratified sampling could be applied to ensure diversity in the
sample, considering factors like age, income, and location.
To propose a plan for random sampling, the business could do the following steps,
 Gather the list or the number of customers who used their services in a
specific period.
 Then, choose a sample or a subset of customers from the total population.
This could be done in a random process.
  Finally, before proceeding, confirm whether the selected sample size is
enough and ensure that the subset represents the entire population to avoid
any errors while using the data.
While applying the stratified sample size to ensure the diversity in the sample,
 While collecting the data, the business should segregate the customers by
their age, location, income, spending level, etc.
 Then divide the different groups into various subsets depending on the
above categories.
 From the total population, these subsets can be used to determine the
percentage of categories that mainly contributed to the business.
 These subsets can be used individually for a study or can be combined by
selecting a few from each subset and creating a different combined data.

c) Consider the following dataset representing the daily sales figures (in dollars) of the store
over a period of one month. The dataset contains 30 data points, each representing the
sales amount for a specific day.

To smooth out the daily sales fluctuations, the store owner decides to calculate a moving
average. A moving average is a statistical technique that involves selecting a specific time
period, summing up the data points within that period, and dividing by the number of
points to obtain an average. This process is then repeated by shifting the time window
one data point at a time until the entire dataset is covered. The resulting moving average
series helps smooth out short-term fluctuations, providing a clearer representation of
underlying trends or patterns in the data. Calculate the 5-day moving average of the
dataset. Show your calculations and your figure.
 337
Mean Absolute Error (MAE) = = 13.48
25

7913
Mean Squared Error (MSE) = = 316.52
25

331.001
Mean Absolute Percentage Error (MAPE) = = 13.24%
25
 SALES FIGURES & CUSTOMER COUNTS
160.00
Actual Daily Data 5 Days Moving Data
150.00

140.00

130.00
VALUE

120.00

110.00

100.00

90.00
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

DAY

d) Explain the trend of sales performance.

 The actual 30 days daily sales value of the shop is between $90 and $150.
 When comparing to the Actual sales performance, the 5 days Moving Average
Forecast shows that,
 At the beginning of 5 days, the 5MA shows a flat sale value as the actual
sale value and from day 6 it starts increasing.
 From day 6 to 12, the 5MA forecast shows an upward trend in sales.
 Days 13, 14, 15 and 16, shows the highest value in the 5MA forecast.
 Days 18 and 25 reflects the lowest value of the 5MA forecast.
 And towards the last week of the month the 5MA forecast shows the
consistent sales value.
 Even while analysing the trend of sales using 5 days moving average forecast, there are
few fluctuations in increase and decrease, hence by recording different months data (more
sample size) and interpreting with 5MA method may gives more accuracy over time.