CASE STUDY: :HYPOTHESIS TESTING
Case study1:
A retail company wants to test whether a new online advertising campaign increases the
average daily sales. Historically, the average daily sales were $10,000. After implementing the
new campaign for one month, they collect sales data and want to analyse if there’s a statistically
significant increase in daily sales.
Objective: To determine whether the new marketing strategy has increased the average daily
sales.
Step 1: Define the Hypotheses
• Null Hypothesis (H₀): The new campaign does not increase average daily sales.
H0: μ=10,000
• Alternate Hypothesis (H₁): The new campaign increases average daily sales.
H1: μ>10,000
This is a one-tailed (right-tailed) test because we're only interested in detecting an increase in
sales.
Step 2: Set the Significance Level
Let’s use a common significance level:
• α=0.05 (This means we're willing to accept a 5% chance of rejecting the null hypothesis
when it’s actually true.)
Step 3: Collect Data and Calculate the p-Value
Assume the company gathers 30 days of sales data post-campaign. The sample mean is $10,500
and the sample standard deviation is $1,200.
Using a one-sample t-test, the p-value is approximately 0.014.
Step 4: Interpret the Results
• p-value = 0.014 and α = 0.02
• Since p-value < α, we reject the null hypothesis.
Conclusion: There is statistically significant evidence at the 5% level to conclude that the
new marketing campaign increased average daily sales.
Ques: When do we reject the NULL hypothesis?
Answer:
If: p-value≤α
Then:
• Reject the null hypothesis (H₀)
• Accept the alternate hypothesis (H₁) as statistically supported (Fail to reject H0)
p-value = the probability of observing the sample results (or more extreme), assuming the null
hypothesis is true.
α (alpha) = your chosen threshold for significance (commonly 0.05, or 5%).
Example
p-value α Decision Interpretation
0.03 0.05 Reject H₀ There’s significant evidence to support H₁.
0.08 0.05 Fail to reject H₀ Not enough evidence to support H₁.
0.049 0.05 Reject H₀ Barely significant; still reject the null.
0.05 0.05 Often reject H₀ Right on the edge; depends on context or convention.
Case study2:
A pharmaceutical company has developed a new drug intended to lower blood pressure. The
average systolic blood pressure for patients with hypertension is known to be 150 mmHg. After
administering the new drug to a sample of patients, the company wants to test if the drug lowers
the average blood pressure.
Step 1: Define the Hypotheses
• Null Hypothesis (H₀): The drug does not lower blood pressure.
H0:μ=150
• Alternate Hypothesis (H₁): The drug lowers blood pressure.
H1:μ<150
This is a left-tailed test because we're testing for a decrease.
Step 2: Choose the Significance Level
• Let’s use α=0.01(very strict, common in medical studies).
Step 3: Collect Data
• After treating 40 patients, the average systolic blood pressure is found to be 145 mmHg,
with a standard deviation of 10 mmHg.
From a t-distribution table or calculator, the p-value is approximately 0.0015.
Step 4: Decision Rule
• p-value = 0.0015
• α = 0.01
Since p-value < α, we reject the null hypothesis.
Conclusion
There is strong statistical evidence at the 1% level to conclude that the new drug significantly
lowers blood pressure.
Component Details
Context Testing if a new drug lowers average systolic blood pressure
Population Mean (μ₀) 150 mmHg (known average for hypertensive patients)
Sample Mean (x̄) 145 mmHg
Sample Size (n) 40 patients
Sample Std. Dev (s) 10 mmHg
Component Details
Test Type One-sample t-test (left-tailed)
Significance Level (α) 0.01 (1%)
Hypothesis Statement
Null Hypothesis (H₀) μ = 150
Alternate Hypothesis (H₁) μ < 150 (drug lowers BP)
Statistic Value
p-value 0.0015 (approx.)
Criteria Result
p-value <= α 0.0015 < 0.01
Conclusion Reject H₀
Interpretation Strong evidence the drug lowers blood pressure
