Hypothesis Testing

Tharindra Nishadini (2304993)

University Canada West

BUSI 650 – Business Analytics

Dr. Ahmed Ali

12th November 2023

 2

Abstract

This report is elaborating the application of hypothesis testing of the statement which

declared by the restaurant manager of a fast-food restaurant by claiming out the average

waiting time for a drive-thru order is 03 minutes. The hypothesis testing has applied for

checking out the significance difference of this claimed statement by fostering the null

hypothesis and alternative hypothesis. The probability value is used to get the decision of

accept or reject of the hypothesis testing with comparison of the significance level. The critical

value also plays a crucial role in hypothesis testing and the comparison between calculated

critical value and z value determines the accept or rejection of the hypothesis testing.

Moreover, the report emphazies the use of hypothesis testing and confidence intervals in the

industries in different situation to make their decisions by reducing the uncertainty of their

decision more reliable way.

 3

Introduction

The fast-food restaurant declared that the average time for a drive thru order is 03

minutes and the restaurant manager requires to claim his observation result with based on the

selected sample and to find out the reliability of the observation result using the statistical data

of z-value, p-value, critical value and the significance level. Moreover, the main objectives of this

report are to find out the null hypothesis and alternative hypothesis with based on the above

statement and build up a consistent 02 hypothesis to get the reliable conclusion at the end.

Further to test the z-value and calculate the probability value based on the z-value and compare

the p-value with significance level of 0.05 and build up a reliable data to reject or accept the

hypothesis secondly. Third objective is to determine the calculated critical value for a one tailed

test with the significance value of 0.05 and compare it with z-value to evaluate the hypothesis

test with rejection or acceptance next objective is find out the influence of when reduce the

sample size by highlighting major drawbacks and introducing the major industries and situations

which hypothesis testing and confidential intervals are applying and how that 02 concepts are

playing major roles in the market to provide reliable and more consistency information which

require in the statistical data analyzing.

Null Hypothesis & Alternative Hypothesis

The null hypothesis is indicating generally no effect or significant difference or

relationship to examine the enough evidence for reject the hypothesis. The alternative

hypothesis is emphasizing the relationship and significant differences by aiming to find out the

enough evidence to support the alternative hypothesis. According to the statement, Null
 4

hypothesis can be stated as, the average time for a drive-thru order is 03 minutes as the fast-

food restaurant declared. It shows as,

H0: μ = 3

Alternative hypothesis is, as stated by fast-food restaurant the average time for a drive-

thru is not 03 minutes by presenting there is a difference in between the declared hours and the

actual hours by restaurant. It represents as,

Ha: μ ≠ 3

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

population standard deviation of 0.4 minutes, calculate the test statistics using a z-value.

The z-test is main component in the hypothesis testing and it emphasizes if there is a

relationship or significant effect, difference between the sample mean and population mean by

making decisions about population details based on the sample details and it highly affect on

the accuracy of the decision. Specially for the market research, quality assurance and etc.

Z-Test equation

Z-value = (x̄ - μ)

( σ / √n)

x̄ : mean wait time of the sample

μ : mean wait time of the population

σ : Standard Deviation of the population

n : Sample size
 5

Calculation

Z-Value = (2.8 – 3.0)

(0.4 /√50)

Z= -3.537

The calculated z-value is -3.537.

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

The p-value is a measurement in statistics which analyzes the proof against the null

hypothesis by showing how to be expected is your observation results occurred by random

probability. The p-value is compared with the significance level of 0.05 and the results can be

indicated as high p-value and low p-value. Low p-value indicates the observation results are

improbable and it leads to reject the null hypothesis. The high p value declares the observed

results is distinctly possible by random chance and there is no powerful evidence to reject the

null hypothesis.

According to this situation, this is a two tailed test since we are finding the significant difference

of the above situation. For an example we are finding, whether the fast-food restaurants’

average time for a drive-thru order is 3 minutes or not and it is directing to the both sides to use

two tailed tests.

By comparing with the standard distribution table, we can find out the probability of

calculated Z-value -3. 537.According to table approximate p-value is 0. 0002.This is a two tailed

test and it should be multiplied by 2 to get the real p-value.

P-value = 2(0.0002)

= 0.0004
 6

The approximate p-value is 0.0004.

Interpretation

According to this situation, the determined significant level is 0.05 and if p-value is less

than or equal to significance level we reject the null hypothesis since low value indicates the

observation result is unlikely.

p ≤ 0.05

if p-value is higher than the significance level we accept the null hypothesis since

observed results is likely by the random chance.

p > 0.05

In this scenario, the probability value is less than the significance level (0.0004 ≤ 0.05)

and the null hypothesis is rejecting. There is powerful evidence to suggest that observation

result is not true, the average time for a drive-thru order is not 03 minutes and the sample data

has provided support to prove the claim made by fast-food restaurant is incorrect. According to

sample data the average waiting time is significantly different from 03 minutes.

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?

Critical value is determined by the based-on significance value and degree of freedom or

distribution of sample. According to this in one-tailed test significant value is 0.05 and we

require to find critical value for comparison with the z-value and it leads to determine whether

to reject the null hypothesis. The critical value can be calculated by using standard distribution

table or statistical calculator.

 7

The distribution is left tailed and the critical value in one-tailed test at a 0.05 significance

level is -1.645 approximately.

 Calculated Z-value : -3.537

 Critical value of one-tailed at a 0.05 significance level : -1.645

Interpretation

Since the calculated z-value farther in the left side of the distribution curve than the

critical value (-1.645), we have powerful evidence to reject the null hypothesis and the

calculation proves that, the average time for a drive-thru order is less than 03 minutes.

In conclude, we reject null hypothesis based on the critical value for a one-tailed test at a

significance level 0.05 this statistic analysis strongly evident that the average wait time for a

drive-thru order is less than 03 minutes by supporting sample data to reject the null hypothesis.

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

Sample size is playing a major role in the statistical analyzing since sample size indicates the

essential factors in an analyzing and directly affect to the conclusion as well. The decreasing of

sample size leads to below affects.

01. Reduce the accuracy of analyzing: The ultimate result or conclusion which you gain from

the smaller sample size may leads to the incorrect directions and the target data with

the smaller sample, your observation result may be father away from the real data with

high margin errors and less certainty.

02. Enhanced the variability: The smaller sample size generates the high inconsistency and

unsteadiness in the data and it become more influential in the analyzing. The results are

making less confidence since the standard deviation creates more influenced by creating
 8

the less representation of the actual standard deviation of population. Essentially, high

variability challenge to acquire the reliable conclusion at the end.

03. Reduce statistical power: The smaller sample size is heading to the high z value and

changes in the p-value. It directly affects to the hypothesis final decision and the

reducing of the sample size leads to decrease the power of statistical analyzing z-value is

influencing by sample size, mean, standard deviation. Thus, with the reducing of sample

size it enhances the higher z-test results (-2.5 Z value) and leads to high extreme test

results as well and the real effect or difference in the population may not explained and

fail to acquire the reliable result at the end.

04. Make exceptional effect on outliers: When sample size is small it is affected greater to

the extreme values in the data set and it creates a deceptive conclusion.

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

confidence intervals could be useful, and briefly explain why.

In the statistics analysis, hypothesis testing and confidence intervals are significantly

guided views for all the researchers and decision makers to analyze their observation results in a

panoramic view. The hypothesis testing is providing the direct validity of the observation results

and confidence intervals provides the range of value that we observe in the sample and for

which we expect to find the value that accurately reflects the population. ( Confidence Intervals -

Statistics Teaching Tools). Together both concepts provide panoramic overview of the statistical

data for all the researchers and some industries are widely using this statistic findings to present

their data more reliable way.

 9

In manufacturing industry quality control is a significant part of the production process

and the hypothesis testing and confidence intervals are playing major role in quality control.

The consistency and stability of the production process of each product is directly affect to the

quality of the each and every product and ensuring the steadiness of production process is

significant in quality control. Thus, the quality controllers in organization hypothesis testing are

widely use for navigate the data on product quality before and after using the new production

process and finding out whether the hypothesis testing is allowing to reject or accept the null

hypothesis and it helps to review if the production process significantly effect to the production

quality. In addition, confidence intervals give the panoramic, boarder perspective to data

analyst for making decisions by showing the uncertainty and instability inherent in the statistical

evaluation and potential strength of the true effect. All manufacturing companies can be used

the confidence interval to assess whether the production process has addressed all the required

product quality characteristics in the process and if not, how they can convert the process in to

successful way to mitigating the risk by creating improvement in the product.

In the investment industry, investors and finance analyst are highly using hypothesis

testing and confidence intervals as a statistical tool to analyze data to making decisions, assess

the particular projects, authenticating presumptions and absorbing more reliable and more

healthy investment opportunities in the market. Investors are using hypothesis basically in

investment analysis to accept or reject the investment assumptions. For an example investor

can assume the particular investment project is making more profit in the market through

hypothesis testing. By analyzing the data related particular investment investors can determine

the observation results is significant or not significant by accepting or rejecting. The interest rate
 10

is the main risk which faced by investors and they can hypothesize whether there is an effect of

interest rate to the investment projects and there is no effect to the investment project. By

analyzing the data relate with statistical analysis investors can be make a judgement regard the

uncertainty and risk management. Moreover, the investors are using confidence intervals in the

analysis of investment project as well. For an example, when calculating the expected return

after a period of time subject to the investment project, the financial analysts have to navigate

the range of possible return on investment with a particular data range and address to the

inborn uncertainty in the market.

In overall, both situations indicate that hypothesis testing and confidence intervals are

playing major roles in the statistical data analysis to acquire the most reliable results at the end.

Conclusion

There is powerful evidence to suggest that observation result is not true, the average

time for a drive-thru order is not 03 minutes and the sample data has provided support to prove

the claim made by fast-food restaurant is incorrect. The probability value is less than the

significance level (0.0004 ≤ 0.05) and the null hypothesis is rejecting. Because according to

sample data the average waiting time is significantly different from 03 minutes. Further we

reject the null hypothesis based on the critical value for a one-tailed test at a significance level

0.05 this statistic analysis strongly evident that the average wait time for a drive-thru order is

less than 03 minutes by supporting sample data to reject the null hypothesis. Moreover, this

report is emphasizing the reducing of sample size effect directly to reduce accuracy of the

analyzing, enhance the variability, reduce statistical power and make exceptional effort on
 11

outliers as well. Finally, the report shows how manufacturing industry and investment market

using data analyzing for their statistical use to obtain the reliable results.
 12

References
Mackowiak, P. A., Wasserman, S.S., and Levine, M.M. (1992). A Critical Appraisal of 98.6
Degrees F, the Upper Limit of the Normal Body Temperature, and Other Legacies of Carl
Reinhold August Wunderlich. Journal of the American Medical Association, 268, 1578-1580.
http://www.stat.yale.edu/Courses/1997-98/101/confint.htm

Confidence Intervals - Statistics Teaching Tools.

https://www.health.ny.gov/diseases/chronic/confint.htm