1 Hypothesis Testing

1.1 Definitions
Definition 1.1. Sampling Distributions: Consider all possible samples of
size n that can be drawn from a given population (with replacement or with-
out replacement). For each sample, we can compute a statistic (such as mean,
standard deviation) that will vary from sample to sample. In this manner we
obtain a distribution of the statistic that is called its sampling distribution.

If, for example, the particular statistic used is the sample mean, then the
distribution is called the sampling distribution of means of sampling distribution
of mean. Similarly, we could have sampling distributions of standard deviations
of medians or proportions etc.
(1) The standard deviation of a sampling distribution of a statistic is often
called its standard error. Standard error for the sampling distributions for
various statistics under the conditions of random sampling from large population
or sampling with replacement from a finite population.

Sampling Distribution Standard Error Special Remarks

Means σX̄ = √σn This is true for large or small samples
q
proportions σP = p(1−p)n This is true for large or small samples

Definition 1.2. Statistical Hypotheses: In attempting to reach decisions,

it is useful to make assumptions (guesses) about the populations involved. Such
assumptions, which may or may not be true, are called statistical hypotheses.

(1) Null and Alternative Hypothesis: According to R.A. Fisher, ”A

null hypothesis is a hypothesis, which is tested for its possible rejection under
the assumption that it is true.” Any other hypothesis which is complementary
to the null hypothesis is called as alternative hypothesis.
Definition 1.3. Errors in Sampling: The main objective of the sampling
theory is to draw valid inferences about the population parameters on the basis
of sample results. In practice, we decide to accept or reject a certain statement
i.e. hypotheses about the population after examining a sample from it. As such,
we liable to commit the following two types of mistakes.
(a) Type I Error: Reject H0 when it is true.
(b) Type I Error: Accept H0 when it is wrong.
In practice, type I error amount to the rejection of a ‘lot’ when it is a good and
type II error may be regarded as accepting a ‘lot’ when it is bad.

Probability of committing mistake of type I is called as level of significance

is is denoted by α. Thus

P (Reject H0 /H0 is correct) = α

1
and also let
P (Accept H0 /H0 is wrong) = β
Definition 1.4. Critical Region: A region corresponding to a static t which
amounts to the rejection of H0 is called as critical region or region of rejection.
If w is the critical region and if t is the value of statistic based on a sample of
size n then
P (t ∈ w/H0 ) = α

1.2 Procedure for Hypothesis Testing (Large Samples)

1- Null Hypothesis:Set up the Null Hypothesis
2- Alternative Hypothesis: Set up the alternative Hypothesis. You may set
up one tailed or two tailed hypothesis depending upon observation from sample.
3- Level of Significance: Choose level of significance α before collecting the
sample.
4- Test Variable: Compute the test statistic or test variable which is generally
given by
t − E.t.(t)
Z=
S.E.(t)
5- Conclusion: We compare the calculated value of test variable with the sig-
nificant value (tabulated value Zα ) at the given level of significance α.

Critical values 1% 5% 10%

Two tailed test Z α2 2.58 1.96 1.645
Right Tailed test (Zα ) 2.33 1.645 1.28
left tailed Test (Zα ) -2.33 -1.645 -1.28
Rules: (a) (Two tailed test): If |Z| ≤ Z α2 then accept the null hypothesis
otherwise reject it.
(b) (Right tailed test): If Z ≥ Zα then reject the null hypothesis otherwise
accept it.
(c) (left tailed test): If Z ≤ −Zα then reject the null hypothesis otherwise accept
it.

1.3 Test of Significance of Mean (Large Samples)

The test variable is
x̄ − µ
Z=
√σ
n

where x̄=sample mean, µ=population mean, σ= population standard deviation,

n=sample size
Example 1.1. A random sample of 400 fishes from a pond has an average
length of 10 cm. Can this be regarded as a sample from a large population with
mean of 10.2 and standard duration of 2.25 cm. (choose α = 1%)

2
Example 1.2. The mean life of sample of 100 aquaria fishes is composed to
be 1570 hours with a standard deviation of 120 hours. The shopkeeper claims
that average life of fishes is 1600 hours. Using the level of significance 0.05 is
the claim acceptable ?
Example 1.3. A sample of 900 members has a mean 3.4 cms is a sample from
a large population of mean 3.25 cms and 2.61 cms. (choose α = 1%
Example 1.4. An insurance agent has claimed that the average age of policy
holders who insure through through him is less than the average of all agent
which is 30.5 yrs. A random sample of 100 policy holders who had insured
through him gave the age distribution.
Age No. of Persons
16-20 12
21-25 22
26-30 20
31-35 30
36-40 16
Calculate the arithmetic mean and standard deviation of this distribution
and use the values to test his claim at level of significance. (Choose α = 5%)

1.4 Test of Significance of Proportions

The test variable is
p̂ − p0
Z=q
p0 (1−p0 )
n

where p̂=sample proportion, p0 =populaton proportion, q0 = 1 − p0 , n=sample

size
Example 1.5. In a sample of 500 women 280 are anemic and the rest are nor-
mal. Can we assume that both normal and anemic are equally is the population.
(choose α = 1% )
Example 1.6. A coin is tossed 900 times and if turns up head 2160 times.
Discuss whether the coin may be unbiased one. (choose α = 5%)
Example 1.7. A certain cubical die was thrown 900 times end a 5 or 6 was
obtained 3240 times on the assumption of random throwing, do the data indicate
an unbiased one. α = 5%)
Example 1.8. In a sample of 1000 peoples in Maharastra, 540 are rice eaters
and the rest are wheat eaters. Can we assume that both rice and wheat eaters
are equally popular in this state at 1% level of significance?
Example 1.9. Twenty people were attacked by a disease and only 18 survived.
Will you reject the hypothesis that survival rate, if attacked by this disease, is
85%, at 5% level of significance.

3
1.5 Test of Significance of Difference between two Means
(Large Samples)
The test variable is
x¯1 − x¯2
Z=q 2
σ1 σ22
n1 + n2

where x¯1 = mean of sample from Population 1, x¯2 = mean of sample from Pop-
ulation 2, µ1 = mean of Population 1, µ2 = mean of Population 2, σ1 = S.D. of
Population 1, σ2 = S.D. of Population 2, n1 = size of sample from Population 1,
n2 = size of sample from Population 2, S1 = S.D. of sample from Population 1,
S2 = S.D. of sample from Population 2.
Example 1.10. In a survey of buying habits, 400 women shoppers are chosen
at random in super market A located in a certain section of the city. Their
average weekly food expenditure is Rs. 250 with a standard deviation of Rs.
40. For 400 women shoppers chosen at random in another supermarket B in
another section of city, the average weekly food expenditure is Rs. 220 Rs. with
a standard deviation of Rs. 55. Test at 1% level of significance whether average
weekly food expenditure of the two populations of shoppers are equal.
Example 1.11. Number of fishes from one pond was taken 1000 and from
second pond 2000. The mean weight of fishes of the two ponds are 67.5 and
68.0 resp. Test the equality of means of the two populations. S.D. for each is
2.5. (Choose α = 5%)

1.6 Test of Significance of Difference between two Means

(Large Samples)
The test Static is
p1 − p2
Z=q
p̂q̂( n11 + n12 )

where p1 = proportion of sample of size n1 from population 1, p2 = proportion

of sample of size n2 from population 2 and p̂ is given by
n1 p1 + n2 p2
p̂ =
n1 + n2
and q̂ = 1 − p̂.
Example 1.12. A random sample of 400 men and 600 women were asked
whether they would like to have a flyover near their residence. 200 men and 325
women were in the favour of the proposal. test the hypothesis that proportions
of men and women in favour of the proposal, are same against that they are
not, at 5% level of significance.
Example 1.13. A company has the head office at Calcutta and a branch at
Bombay. The personnel director wanted to know if the workers at the two places

4
would like introduction of a new plan of work and a survey was conducted for
this purpose. Out of sample of 500 workers at Calcutta, 62% favoured the new
plan. At Bombay out of a sample of 400 workers, 41% workers were against the
new plan. Is there any significant difference between the two groups in their
attitude towards the new plan at 5% level?
Example 1.14. Before an increase in excise duty on tea, 800 percent out of a
sample of 1000 persons were found to be tea drinkers. After the increase in excise
duty, 800 people were tea drinkers in a sample of 1200 peoples. State whether
there is a significant decrease in the consumption of tea after the increase in
excise duty? (at 5% level of significance)

2 Small Sample Test

Sometimes it is not possible to have large samples on account of economical or
any other factors so we depend on small samples. The entire theory of hypothesis
testing for large samples, is based on Normal Tests. However, if the sample size
is small, the distribution of various test statistic is not normal one and in such
cases normal test cannot be applied. Therefor, we use exact sampling tests
introduced by W.S. Gosset who wrote under the pseudo name ‘Student’. Here,
we shall consider following methods, known as (1) t-test and (ii) F-test. The
small sample tests can be applied to large samples but the converse is not true.
In all exact sample tests we assume that the parent population is normally
distributed.

2.1 Student’s t and t-distribution

W.S. Gosset was the first to define the distribution of statistic t. Let {x1 , x2 , ..., xn
be the members of a sample of size n drawn from a normal population with mean
µ and variance σ 2 . Then the following statistic (called as Student’s t sometimes
as t-statistic)
x̄ − µ
t= s
√
n
Pn 2
i=1 (xi −µ)
Pn
where x̄ = i=1 xi is the sample mean and s2 = (n−1) . If we take the
different samples the distribution of t is given by
y0
y= t2 (ν+1)/2
(1 + ν )

where y0 is a constant depending on n and ν = n−1 is called degrees of freedom.

2.2 Test of Significance of Mean (Small Samples, t-test)

The test variable is
x̄ − µ
t=
√s
n

5
 q Pn 2
i=1 (xi −µ)
where x̄=sample mean, µ=population mean, s = (n−1) sample stan-
dard deviation, n=sample size. if the population standard deviation is known
then use
x̄ − µ
t= σ
√
n−1

Example 2.1. A mechanic is making certain engine parts with the diameter
0.700 inches. A random sample of 10 parts shows that a mean diameter of 0.742
inch with a standard deviation of 0.040 inch. Compute the test statistic you
would like to test. Whether the work is meeting specifications.
Example 2.2. The mean weekly sales of a certain bathing soap in departmental
stores was 146.3 bars per store. After an advertising campaign the mean weekly
sales in 22 stores for a typical week increased to 153.7 and showed a standard
deviation of 17.2. Was the advertising campaign successful?
Example 2.3. Ten individuals are chosen at random from a population of and
their heights are found to be in inches 68, 63, 64, 65, 66, 69, 69, 70, 70, 71.
Discuss the proposal that mean height in the universe is 65 inches given that
for 9 degrees of freedom Student’s t at 5% level of significance is 2.262.
Example 2.4. A random sample of 10 boys had the following I.Q.’s: 70, 120,
110, 101, 88, 83, 95, 98, 107, 100. Do these data support the assumption of that
population mean I.Q. is 100. (choose 5% level.)