MA2140/MA20120: Introduction to Statistics

Dr. Sameen Naqvi

Department of Mathematics, IIT Hyderabad
Email id: [email protected]
 Some important terms

I Test statistic: The sample statistic one uses to either reject

H0 (and conclude Ha ) or not to reject H0 .

I Rejection Region: The set of values for the test statistic

that leads to rejection of H0 .

I Non-Rejection Region: The set of values not in the

rejection region that leads to non-rejection of H0 .

I Critical values: The values of the test statistic that separate

the rejection and non-rejection regions.
Rejection region approach for making a statistical decision
 Rejection region approach

I Step 1: Check the conditions necessary to run the selected

test and select the hypotheses for that test.

I Step 2: Decide on the significance level, α.

I Step 3: Compute the value of the test statistic.

I Step 4: Find the appropriate critical values for the tests and
write down clearly the rejection region for the problem.

I Step 5: Check to see if value of the test-statistic falls in the

rejection region. If it does, then reject H0 (and conclude Ha ).
If it does not fall in the rejection region, do not reject H0 .

I Step 6: State the conclusion in words.

Tests about One mean

1. When σ is known
 Case 1: Right-tailed test

Hypotheses:

H0 : µ ≤ µ 0 and Ha : µ > µ0

TS:
X − µ0
Z∗ = √
σ/ n

RR: Reject H0 if Z ∗ ≥ Zα .
 Case 2: Left-tailed test

Hypotheses:

H0 : µ ≥ µ 0 and Ha : µ < µ0

TS:
X − µ0
Z∗ = √
σ/ n

RR: Reject H0 if Z ∗ ≤ −Zα .

 Case 3: Two-tailed test

Hypotheses:

H 0 : µ = µ0 and Ha : µ 6= µ0

TS:
X − µ0
Z∗ = √
σ/ n

RR: Reject H0 if |Z ∗ | ≥ Zα/2 .

 Example 1

I Suppose a smartphone manufacturing company wishes to

determine if their products typically last longer than 40
months. It’s assumed that the lifespan of these smartphones
follows a normal distribution, and historically, the standard
deviation has been around 15 months.

I To investigate this, the company selects a sample of 50

smartphones and records their lifespans. The sample reveals
an average lifespan of 45 months.

I Is their significant evidence to support the claim that the

average smartphone lifespan exceeds 40 months? Take
α = 0.05.
 Solution

I Step 1: We can use the one-mean Z-test and set up the

research hypothesis:
H0 : µ ≤ 40 vs Ha : µ > 40

I Step 2: Here, α = 0.05.

I Step 3: Compute the value of test statistic:

x − µ0 45 − 40
Z∗ = √ = √ = 2.357
σ/ n 15/ 50
 Solution contd.

I Step 4: From the table, we find that Z0.05 = 1.645, and thus,
the critical value is 1.645.

I Step 5: Since the value of the test statistic falls in the

rejection region (i.e., Z ∗ ≥ 1.645), we reject H0 .

I This indicates that there is sufficient evidence to support the

claim that the average lifetime of the smartphones is greater
than 40 months at 0.05 significance level.
2. When σ is unknown
 Case 1: Right-tailed test

Hypotheses:

H0 : µ ≤ µ 0 and Ha : µ > µ0

TS:
X − µ0
t∗ = √
S/ n

RR: Reject H0 if t ∗ ≥ tα .
 Case 2: Left-tailed test

Hypotheses:

H0 : µ ≥ µ 0 and Ha : µ < µ0

TS:
X − µ0
t∗ = √
S/ n

RR: Reject H0 if t ∗ ≤ −tα .

 Case 3: Two-tailed test

Hypotheses:

H 0 : µ = µ0 and Ha : µ 6= µ0

TS:
X − µ0
t∗ = √
S/ n

RR: Reject H0 if |t ∗ | ≥ tα/2 .

 Example 2

I Imagine that Hyundai Motor Company is curious about the

average fuel efficiency of their cars, with the company’s
manager asserting that a car should average 25 km/liter.
Assuming the mileage follows a normal distribution, the
company decides to verify this claim.

I To do so, they examine 40 of their vehicles and find that the

average fuel efficiency is 28 km/liter, and the standard
deviation across these vehicles is 10 km/liter.

I At α = 0.01, determine if there is enough statistical evidence

to conclude that the average mileage is different from the
manager’s claim.
 Solution

I Step 1: We can use the one-mean t-test and set up the

research hypothesis:
H0 : µ = 25 vs Ha : µ 6= 25

I Step 2: Here, α = 0.01.

I Step 3: Compute the value of test statistic:

x − µ0 28 − 25
t∗ = √ = √ = 1.8973
S/ n 10/ 40
 Solution contd.

I Step 4: From the table, we find that t0.005 = 2.805, and thus,
the critical value is 2.023.
The rejection region for the two-tailed test is given by:

t ∗ < −2.023, or t ∗ > 2.023.

I Step 5: Since the value of the test statistic (t ∗ ) doesn’t fall

in the rejection region, we fail to reject H0 .

I Hence, we do not have enough statistical evidence to reject

the manager’s claim that the average mileage of the cars is
25km/liter .
Test for One Proportion
 Case 1: Right-tailed test
Hypotheses:

H0 : p ≤ p 0 and Ha : p > p 0

TS:
p̂ − p0
Z∗ =
σp̂

RR: Reject H0 if Z ∗ > Zα .

I Note: Under H0 ,
r
p0 (1 − p0 )
σp̂ =
n

I Also, n must satisfy np0 ≥ 5 and n(1 − p0 ) ≥ 5.

 Case 2: Left-tailed test

Hypotheses:

H0 : p ≥ p 0 and Ha : p < p 0

TS:
p̂ − p0
Z∗ =
σp̂

RR: Reject H0 if Z ∗ < −Zα .

 Case 3: Two-tailed test

Hypotheses:

H0 : p = p 0 and Ha : p 6= p0

TS:
p̂ − p0
Z∗ =
σp̂

RR: Reject H0 if |Z ∗ | > Zα/2 .

 Example 3

A national fitness survey recently found that 44% of adults

participate in regular exercise, defined as at least 150 minutes
of moderate-intensity activity per week. The mayor of a
health-conscious small town claims that their community is
more active than the national average, boasting a population
of 25,000 adults.
To investigate the mayor’s assertion, the local health
department surveyed a random sample of 2,500 adults from
the town. They discovered that 1,200 of these adults met the
criteria for regular exercise according to the survey’s standards.
Determine whether there is substantial evidence to support
the mayor’s claim that a greater percentage of their adult
population engages in regular exercise compared to the
national figure, with a significance level set at 5%.
 Solution
I Let p be the proportion of adults in the town who engage in
regular exercise.
I Step 1: We can use the Z-test and set up the research
hypothesis:
H0 : p ≤ 0.44 vs Ha : p > 0.44
I Step 2: Here, α = 0.05.
I Step 3: Compute the value of test statistic:

p̂ − p 0.48 − 0.44
Z∗ = = = 4.03
σp̂ 0.00993
where
r r
p0 (1 − p0 ) 0.44(1 − 0.44)
σp̂ = = = 0.00993.
n 2500
 Solution contd.

I Step 4: For α = 0.05, we find that Z0.05 = 1.645, and thus,

the critical value is 1.645.

I Step 5: Since the test statistic is greater than the critical

Z-value (4.03 > 1.645), we reject the null hypothesis.

I There is sufficient evidence to support the claim that a higher

percentage of adults in the town engage in regular exercise
compared to the national figure.
Thank you for listening!