A level statistics teaching notes Prepared by Manyuvire D Cell : 0783235483

ESTIMATION AND SAMPLING

CENTRAL LIMIT THEOREM AND CONFIDENCE INTERVALS

SYLLABUS OBJECTIVES (6046)

 state the Central Limit Theorem

 recognize that the sample mean can be regarded as a random variable
 use the Central Limit Theorem
 identify the implications of the Central Limit Theorem on small and large samples
 determine a confidence interval for a population mean in cases where the population
is normally distributed with known variance or where a large sample with unknown
variance is used
 determine a confidence interval for a population mean in cases where the population
is normally distributed with unknown variance where a small sample is used
 determine from a large sample an approximate confidence interval for a population
proportion

SAMPLE MEAN AS A RANDOM VARIABLE

 Samples of same sizes can be repeatedly taken at random from a population and the
sample mean for each calculated.

 When sample means take a range of values the sample mean will become a random
variable.

DISTRIBUTION OF SAMPLE MEANS

 It is a probability distribution of means calculated from all possible random samples

of a specific size taken from a population.

EXAMPLLE
1. A man makes three sizes of toasters. 40% of the toasters sell for $16, 50% sell for
$20 and 10% sell for $30. A sample of two toasters is send to a shop.
a) List all possible prices of the samples that could be sent.
b) Find the distribution of the sample means.

1
A level statistics teaching notes Prepared by Manyuvire D Cell : 0783235483

Solution
a) (30 30) (20 20) (16 16) (16 20) (20 16) (16 30) (30 16) (20 30) (30 20)
b) The means for the samples in a) are 30 20 16 18 18 23 23 25 25 respectively
 P(30 and 30) = P(16 and 30) =
P(20 and 20) = P(30 and 16) =
P(16and16) = P(20 and 30) =
P(16 and 20) = P(30 and 20) =
P(20and16) =
The distribution of means therefore is:

16 18 20 23 25 30

P(

THE DISTRIBUTION OF WHEN X IS FROM A NORMAL POPULATION.

A random variable X1 X2 X3 … Xn is taken from a normal population X N( , )

If X is a random variable X1+ X2 +X3 +… +Xn then = (X1+ X2 +X3 +… +Xn )
E =E{ (X1+ X2 +X3 +… +Xn )}
= (E(X1) +E( X2 )+E(X3 )+… +E(Xn ))
= ( 1+ 2 + 3 +… + n )
=
=
Therefore the mean of the sample means is the same as the population mean. ie E =

Var(X) = Var{ (X1+ X2 +X3 +… +Xn )}

= (Var(X1) +Var( X2 )+Var(X3 )+… +Var(Xn ))
= ( 1+ 2 + 3 +… + n )

Therefore if X N( , ) then N

2
A level statistics teaching notes Prepared by Manyuvire D Cell : 0783235483

CENTRAL LIMIT THEOREM

 If samples are taken from a population that is not normally distributed, the sampling
distribution will take the characteristics shape of a normal distribution as the sample
sizes increase. As n gets large is approximately normal.

 The Central limit theorem is that: For a sample taken from a non normal population
with mean and variance , by central limit theorem N provided
.

EXAMPLES
1. A sample of size 9 is taken from a population with distribution N(10 ; 22) Find the
probability that the sample mean ( ) is more than 11.
Solution
X N( , ) then N =N

P( =P

= P(Z > 1,5)

= 1 ─ 0,9332
= 0,0668
2. A tetrahedral with four sides marked 1,2,3,4 is rolled 50 times and the mean score
of the 50 scores is recorded. Let X be the random variable the score on a single roll
with the following distribution:

X 1 2 3 4

P(X
=x)

Find the probability that the sample mean is over 2.

Solution
E(X) = = =2,42
E(X2) = = = 6,92
Var(X) = = = 1,08

By the central limit theorem N =N

P(P( =P

= P(Z >─2,84)
= 1 ─ (1─0,9977)
= 0,9977

3
A level statistics teaching notes Prepared by Manyuvire D Cell : 0783235483

3.Forty observations are taken from a population with a distribution given by the probability

density function

a) Find mean and variance.

b) Find the probability that the sample mean is greater than 2,10
Solution
E(X) = dx = = ─0=2

E(X2) = dx = = ─ 0 = 4,5
Var(X) = 4,5 ─ 22 = 0,5
Therefore then N =N

P( >2,10) = P

= P(Z >─0,8944)
= 1 ─ (1─ )
= 0,8144
4.A sample of n independent observations is taken from a normal population with mean 74
and standard deviation 6. The sample mean is denoted by .
a) Find n is P( = 0,282 b) ) Find n is P( = 0,0037
Solution
N

P( = 0,282 P =0,282

P = 0,718

= 0,577
= 3,462
n = 11,985
n = 12

P( = 0,282 P =0,0037

= (0,0037) = (0,9963) by symmetry

= ─2,68

4
A level statistics teaching notes Prepared by Manyuvire D Cell : 0783235483

=─16,08
=
=4,466666667
n= 19,95
n = 20

CONFIDENCE INTREVAL

 It is a range of values that is believed to include an unknown population parameter.

 It is a range of values where the probability of including an unknown population
parameter is known. This probability is called confidence level.
 For example 95% is the probability that a population parameter is included in the
interval a to b can be illustrated below.

- is the probability that it is not contained and this is 5 % and is equally the area
under both sides of the distribution ie 2,5% on each as shown. We need to find the Z
values for a and b ( ) and these are −1,96 and 1,96 for a and b
respectively.
- P( −1,96<Z<1,96) = 0,95
- By Central limit Theorem N( and Z =

- P( −1,96<Z<1,96) = 0,95 P( −1,96 < < 1,96)

−1,96 × < < 1,96×

−1,96 × – − 1,96× − (Isolate )
1,96 × + −1,96× + Divide by negative 1
−1,96 × +1,96×
1,96×
Therefore 95% Confidence Interval is given by 1,96×
Generally Confidence interval is given by ×

5
A level statistics teaching notes Prepared by Manyuvire D Cell : 0783235483

The formulae for Confidence interval depends on the sample size and whether the
is known or not.
Known variance , unknown Unknown variance and n 30
variance and n 30
× ×

EXAMPLES
1. A random sample of size 9 is taken from a normal distribution with variance 36
and sample mean 128. Find a 99 % and 95% confidence intervals for the population
mean.
Solution
For 95%, = = 1,96 For 99%, = = 2,576
95% CI = × 99% CI = ×

= 128 ±1,96 = 128 ±2,576

= 128 ± 1,96(2) = 128 ± 2,576(2)
=128 ± 3,92 =128 ± 5,156
= [124,04 131,92] = [122,844 133,156]
= [123 ; 133] to 3 s.f

2. A normal distribution has mean and variance 4,41. A random sample has the
following values. 23.1, 21.8, 24.6, 22.5 . Use this sample to find 98% comfidence
interval for the mean.
Solution
= = = 23
For 98%, = = 1,282
98% CI = ×

= 23 ±2,326
= 23 ± 2,4423
= [20,5577 25,4423]
= [20,6 25,4] to 3 s.f

3. During a particular evening, 10 babies were born on a particular maternity ward in

a large hospital. The lenghts, in centimeters, of the babies were noted
50, 51, 45, 47, 49, 48, 54, 53, 45, 50. Assuming that the sample came from an
underlying normal population, calculate 95% confidence interval for the mean of the
population.
Solution

= = = 49,2 = = = .

6
A level statistics teaching notes Prepared by Manyuvire D Cell : 0783235483

Since n is small and was estimated we use ×

= 2,262

95% CI = 49,2 ± 2,262×

= 49,2 ± 2,262×

= 49,2 ± 2,18
=[47,02 ; 51,38]

4. The 95% confidence interval for the mean life of light bulbs constructed from a
sample of size 36 is (1023,3hrs; 1161,7hrs) Assuming that the life of light bulbs is
normally distributed find the 99% confidence interval for the mean life of this brand
of bulbs.
Solution
[1023,3 ; 1161,7] = × ; ×
[1023,3 ; 1161,7] = × ; ×
× = 1023,3 ------------ (1) and × = 1161,7 ---------- (2)
Solving (1) and (2) simultaneously we get = 1092,5 and = 211,84
99% CI = ×
= 90,9
=[1001,6 ; 1183,4]

5. A 95% confidence interval for a mean is 85,3 ± 2,35. Find the 99% confidence
interval for .
Solution
For 95%, = = 1,96
For 99%, = = 2,576
× = 2,35
99% CI = 85,3 ± 2,576 ×
=
= 85,3 ± 3,09

WIDTH OF CONFIDENCE INTERVAL

 This is the difference between the upper confidence limit and the lower
confidence limit.
 For a confidence interval × which is × , × ,
the difference between the upper limit and the lower limit is given by :
× – × )

7
A level statistics teaching notes Prepared by Manyuvire D Cell : 0783235483

= × – ×

=2

 Generally width = 2 For Known and unknown , with n 30 and

2 when is Unknown and n 30

EXAMPLES
1. A random sample of size 9 is taken from a normal distribution with variance 36
and sample mean 128. Find the width of 99 % and 95% confidence intervals for the
population mean.
Solution
For 95%, = = 1,96 For 99%, = = 2,576

× = 1,96 = 3,92 × = 2,576 = 5,156

Width = 2 Width = 2
= 2(3,92) = 2(5,156)
= 7,84 = 10,312

SAMPLE SIZE

 The sample size n depends on the margin of error, population standard deviation and
the degree of confidence. The margin of error depends on the bound on the error e
which is the value × on the confidence interval ×

-e= × e= e =

 Generally sample size n for the sample mean to be within e of the true mean is given
by

EXAMPLE
1. A normal distribution has standard deviation 15. Estimate the sample size required
for one to be 90% confidence that the sample mean will be within 0,5 of the true
mean.

8
A level statistics teaching notes Prepared by Manyuvire D Cell : 0783235483

Solution
× = 0.8
n=
1,645× =0,8
=
=
= 951,3369141
= 951 =
= 951,3369141
OR = 951
× = ± 0.8

SAMPLE SIZE FOR A SPECIFIED WIDTH

 Use the formula for calculating width and make an inequality or equation
involving it and the specified width value.

EXAMPLES
1. A normal distribution has standard deviation 15. Estimate the sample size
required if 90% confidence interval is required for the mean to have a width of
less than 2
Solution
For 90% CI = = 1,645

2(1,645 ×

608,855625 <
= 609

CONFIDENCE LEVEL WHEN GIVEN WIDTH

Example
A random sample of size 9 is taken from a normal distribution with variance 36 and
sample mean 128. What Confidence level will be associated with a width of 7,84
Solution
A width of 7,84 7,84 =2

7,84 =2
7,84 = ×4
= 1,96

9
A level statistics teaching notes Prepared by Manyuvire D Cell : 0783235483

Confidence level is the area between ± 1,96

= 1─ [P(Z>1,96)+P(Z<─1,96)]
= 1─[1─ +(1─ ]
= 1─[2─2 ]
= 1─[2─ 2(0,975)]
=1─0,05
=0,95
=95%

CONFIDENCE INTREVAL FOR A POPULATION PROPORTION

 The confidence interval for a particular proportion p is based on the sampling
distribution of .
 The confidence interval must meet the criteria that and so as
to use the normal approximation to the binomial.
 The sampling distributions of proportions is approximately normal provided n
is large such that N where q = 1─p

 The confidence interval for a population proportion is given by ±

EXAMPLE
1. In a survey of a random sample of 250 households in a large city, 170
households owned at least one pet. Find an approximate 95% confidence
interval for the proportion
Solution
= = and =
For 95% = = 1,96

Confidence interval = ±

= ±
= 0,68±0,058
= [0,622 ; 0,738]

2. A survey of 225 families found that 25% of the families purchase meat at
least once a week. Find the 95% confidence interval of the true proportion of
people who purchase meat at least once a week.
Solution
= = and =

10
A level statistics teaching notes Prepared by Manyuvire D Cell : 0783235483

For 95% = = 1,96

Confidence interval = ± = ±
= 0,25±0,057
= [0,193; 0,307]

ESTIMATING SAMPLE SIZE FOR PROPORTION

 We may wish to determine how large a sample is necessary to ensure that the error in
estimating μ will not exceed e - the ‘bound on the error’

 From ± , e=

Making n the subject, = =

= n= =

 The minimum sample size required to estimate the population proportion to be within

a specified amount e with a specific level of confidence is given by: n =

EXAMPLE
1. A researcher wishes to estimate, with 95% confidence, the proportion of people
who purchase meat at least once a week. A previous study shows that 25% of those
interviewed purchase meet once a week. Calculate the minimum sample size required
for the researcher to be within 5% of the true proportion.
Solution
For 95%, = = 1,96
= = and =

n=

n=
n = 288,12
n= 288

COMPILED BY MANYUVIRE D

COMMENTS ON ERRORS AND OMMISSIONS ARE WELCOME

YOU CAN CALL TEXT OR WHATSAP 0783 235 483

ONLINE LESSONS ALSO OFFERED

11