A LEVEL STATISTICS TEACHING NOTES BY MANYUVIRE D CELL 0783235483

SAMPLING AND ESTIMATION

SAMPLING TECHNIQUES AND POINT ESTIMATIONS
SYLLABUS OBJECTIVES (6046)

 distinguish between sample and population

 distinguish between probability sampling techniques and non probability sampling

techniques

 apply the sampling methods to identify representative samples

 calculate sample mean, variance and standard deviation

 find the unbiased estimates of population parameters

 solve problems involving sampling and estimation

1. POPULATION AND SAMPLE

 Population refers to the collection of all elements about which
information is sought. For example all A level students in Zaka
District.
 Sample refers to a sub group /subset of a population containing
elements about which information is sought. For example 100 A
level students in Zaka District.

2. PARAMETER AND STATSTIC

 Parameter : is a calculation from a population for example
population mean and variance 2
 Statistic: is a calculation from a sample for example sample mean
and sample variance s2

1
A LEVEL STATISTICS TEACHING NOTES BY MANYUVIRE D CELL 0783235483

3. DIFFERENCES BETWEEN POPULATION AND SAMPLE

POPULATION SAMPLE
-Is large since it constitutes all the -Is small since it is a subset of a
elements under study population
-calculations from a population are -A calculation from a sample is
called parameters called a statistic

4. When carrying out researches it is more convenient to use samples

instead of populations.
 When the investigation involves the entire population is used it is
called a census and when an investigation involves a sample it is
called a sample survey.
 COMPARISONS BETWEEN POPULATION CENSUS AND SAMPLE
SURVEY
POPULATION CENSUS SAMPLE SURVEY
Advantages Advantages
-Gives reliable information since -Saves time
all members are investigated - Saves money
-Information obtained is -Suitable for investigations that
accurate result in destruction of
-It is unbiased elements in the population.
Disadvantages Disadvantages
-Time consuming -Is not free from bias.
-Very expensive to carry out -Elements in the sample may
-Not suitable for investigations not be representative of the
that result in destruction of population
elements in the population. - Information obtained may not
be reliable.

5. Samples should be representative of the population from which they are

obtained. Random samples therefore should be used.
 A random sample is a sample where every possible sample of size
n has an equal chance of being chosen. It is unbiased.

6. SAMPLING TECHNIQUES
 Different methods of sampling can be used to come up with a
representative sample.

2
A LEVEL STATISTICS TEACHING NOTES BY MANYUVIRE D CELL 0783235483

 There are two groups of methods which are probability sampling

techniques and non probability sampling techniques.

7. PROBABILITY/ RANDOM SAMPLING METHODS

 These are method where every element of a population has a
known or equal probability of being selected.
 They entail choosing of elements by chance from a sampling
frame.
 A sampling frame is a list of all elements in the population.

7 a) SIMPLE RANDOM SAMPLING

 Gives rise to a random sample.

 Can be done in two ways which are with replacement and without
replacement.
 With replacement entails that every element has an equal chance of
being selected. A member chosen at each draw is replaced before the
next draw is done. This is suitable for a small population.
 Without replacement entails that members will not have equal chances
of being selected. A selected member is not replaced. This is suitable for
a large population.
 Two commonly methods of simple random sampling are drawing
lots/hat system and random number sampling.

DRAWING LOTS/HAT SYSTEM

1 Have a sampling frame by allocating numbers to the population.
2. Make identical cards written numbers for each element in the
population.
3. Put the numbered cards in a bowel or hat then shake it to randomly
mix the cards.
4. Pick or fish out cards randomly from the bowel or hat one at each
draw. This can be done with or without replacement depending on the
size of the population. Repeat draws until you reach the required
number in a sample.

3
A LEVEL STATISTICS TEACHING NOTES BY MANYUVIRE D CELL 0783235483

ADVANTAGES DISADVANTAGES
-Cheap to carry out -not suitable for very large
-Simple to do populations
-it is random -A sampling frame is required

RANDOM NUMBERS
This can be done using random number tables or using a calculator
Example 1
Using a given extract of random numbers select a sample of 10 students
from a population of 50 students.

65 23 68 00 77 82 58 14 10 85 11 35
09 56 76 51 04 73 94 30 16 74 69 59
55 99 98 60 01 33 06 93 85 13 23 17
72 82 45 44 09 53 04 83 03 83 98 41
04 21 28 72 73 25 02 74 35 81 78 49

SOLUTION
Assign numbers 00; 01; 02; . . .; 48; 49 to the elements in the population.(ie
make sampling frame)
Start from the top left and work to the right. When you reach the last digit to
your right, start again from the left with the next unused digit along the first
column. The sample will have the elements allocated (23; 00; 14; 10; 11; 35; 09;
04; 30; 16)
NB You can start at any randomly selected place and travel in any direction of
your choice.

Example 2

Use a given extract of random numbers from a calculator random number

generator to select the first five digits of a sample from a population of 50
students.
0.165; 0.186; 0.128; 0.308; 0.881; 0.285; 0.585; 0.774; 0.978; 0.776; 0.745;
0.392; 0.17; 0.851

4
A LEVEL STATISTICS TEACHING NOTES BY MANYUVIRE D CELL 0783235483

SOLUTION
Assign numbers 00; 01; 02; . . .; 48; 49 to the elements in the population.(ie
make sampling frame)
There are various ways of interpreting the numbers for example ignoring the
decimals and taking the first two digits or the last two. Or using digits after the
comma and grouping them in twos.
Using the last option the numbers will be like
16 51 86 12 83 08 88 12 85 58 57 74 97 87 76 74 53 92 17 85
the first 4 of the sample working from left to right are 16; 12; 08; 12; 17
NB Ignore repeats.

ADVANTAGES DISADVANTAGES
-Numbers are truly -not suitable for very large
random and far from populations
bias -A sampling frame is required
-Easy to use
-Each number has an
equal chance to be
selected

7b) SYSTEMATIC SAMPLING

 It involves choosing elements at regular intervals from an ordered list.

 A kth element can be taken from the sampling frame where k =

Example 3
Select a sample of 10 students from 50 students.

SOLUTION
1.have a sampling frame by assigning students numbers 1 up to 50
2. Calculate k ie k = = 5th element
3. Pick a random number to start from. For example a number between 1 and 5
say the number is 2. The net number will be the 5th number after 2 ie 7 the next
will be 5th after 7 ie 12 and so on.
4. The sample will have the students numbered
(2; 7 12; 17; 22; 27; 32; 37; 42; 47)
5
A LEVEL STATISTICS TEACHING NOTES BY MANYUVIRE D CELL 0783235483

NB Select the first number randomly

The method is suitable when the population is too large for simple random
sampling.

ADVANTAGES DISADVANTAGES
-Simple to use -Can introduce bias.
-Suitable for large -A sampling frame is required
populations

7c) STRATIFIED RANDOM SAMPLING

 It is a form of random sampling in which the population is divided into

sub groups called strata.
 The strata can be for example, age, religion, gender etc.
 A simple random sample is selected within each stratum.
 Proportional representation in strata should be used. Number in each
stratum is calculated as

Example 4
A population of students who participate in different sports disciplines
where asked about their Sports director and teachers were treating
them during sporting activities. 150 students constitute the population.
Three sporting disciplines were soccer, volleyball and Hand ball and the
number in each discipline were 80, 40 and 30 respectively. A sample of
30 is to be selected. How many students will be selected per each
discipline?

SOLUTION
soccer = = 16 volleyball = = 8 and Handball =
=6
The elements from each stratum are then selected randomly.
ADVANTAGES DISADVANTAGES
-Reflects population -May overlap if strata are not
structure clearly defined.
-Gives accurate information -A sampling frame is required for
where strata are clear. each stratum

6
A LEVEL STATISTICS TEACHING NOTES BY MANYUVIRE D CELL 0783235483

7d) CLUSTER

 Is a method used where population consist of homogeneous sub groups

and within those groups the elements exhibit different characteristics.
 Pre existing or natural groups called clusters are selected to be included
in the sample.

ADVANTAGES DISADVANTAGES
-Reduce costs -Elements may not have same
-Convenient variations in characteristics as
-Less time consuming elements selected from
population.
-Some sections of the population
may be under or over
represented.

8. NON PROBABILITY/NON RANDOM SAMPLING TECHNIQUES

 These are methods that are non probabilistic in nature.
 They are used where it is not possible to use to use random
methods and where a sampling frame is difficult to come by.

8 a) QUOTA SAMPLING

 It is a method sued where the population can be divided into strata but
where the sampling frames cannot be obtained. It is similar to stratified
in that you start by identifying the strata in the population and divide
accordingly.
 Quotas are specified first. For example How many males or females to
include. It will be up to the interviewer to choose who is sampled once
the quota is specified.
 The method is suitable for opinion surveys.

ADVANTAGES DISADVANTAGES
-Cheap and easy to use -non random and susceptible to
since no sampling frame bias.
is required - non – respondents are not
recorded

7
A LEVEL STATISTICS TEACHING NOTES BY MANYUVIRE D CELL 0783235483

8b) CONVINIENCE SAMPLING

 This is a method when a researcher selects those elements which are

easily accessible to him.
ADVANTAGES DISADVANTAGES
-Simple and cheap -Can introduce bias.
- Sample obtained is not
representative of the population

9. SAMPLE STATISTIC
 A Statistic is a quantity calculated sorely from the observation in a
sample. These include sample mean , sample variance s2 and sample
standard deviation s.
 = s2 = or

Example 5
The following is a sample 10 families showing of number of children in a
family taken from a village population. 5 4 3 4 7 1 2 4 6 5.
Calculate a) sample mean b) sample variance c) sample standard
deviation.

SOLUTION
a) = = = = 4,1
b)
x 5 4 3 4 7 1 2 4 6 5 = 41
x2 25 16 9 16 49 1 4 16 36 25 = 197

s2 =

=
= 3,21 to 2 d.p

s= = 1,79 to 2 d.p
Or the data can be entered into the calculator in Statistic SD mode
Enter mode 1-0 (using sharp EL-531WH)

8
A LEVEL STATISTICS TEACHING NOTES BY MANYUVIRE D CELL 0783235483

To enter 5, press 5 then DATA. Do the same to all values until all data is
entered.
You can now retrieve your statistics using RCL. For example for sample
standard deviation RCL = 1,79 RCL =4,1
NB Consult your manual for other calculators.
10.UNBIASED ESTIMATES OF POPULATION PARAMETERS
 It involves the use of sample statistics to predict corresponding
population parameters. A sample statistic will be an unbiased
estimator of a population parameter.
 Estimate of = estimate of = and estimate of p =

POPULATION UNBIASED ESTIMATE

PARAMETER
Population mean = =
Population variance =
2
Or = where = =
ie sample variance
Population proportion = where is sample proportion.
p

Example 6

Cartons of oranges are filled by a machine. A sample of 10 cartons selected at

random from the population contained the following quantities in millilitres.
201,2 205,0 2091 202,3 204,6 206,4 210,1 201,9 203,7 207,3 Calculate the
unbiased estimate of variance of the population from which the sample was
taken.
SOLUTION
= = = = 205,16

= = = 9,223

Example 7
Find the unbiased estimate of the population mean and variance of the
population from which the samples with the following information are drawn.

9
A LEVEL STATISTICS TEACHING NOTES BY MANYUVIRE D CELL 0783235483

a) = 120 = 2102 n = 8
b) = 5472 = 1236 n = 12
c) = 270 = 2540 n = 100

SOLUTIONS
a) = = = 15 = = = 43,142
b) = = = 456
= 1236
= 1236
− 912 + )= 1236
– 912 + 207936(12) = 1236
= 1236 + 912 − 207936(12)
= 2496468
= = = 112,36

c) = 270
− ) = 270
− = 270
= 5270
=
= = 52,70

For variance find

= 2540 = 2540
− 100 + )= 2540
– 100 + 2500(100) = 2540
= 2540 + 100 − 2500(100)
=279540
= = = 18,29

OR Let Y be the random variable x− 50 = 270 E(Y) = = 2,70

E(x−50) = E(X) −50

10
A LEVEL STATISTICS TEACHING NOTES BY MANYUVIRE D CELL 0783235483

E(x−50) = E(Y) = E(X) – 50 = 2,70

E(X) – 50 = 2,70
E(X) = 2,70 + 50 = 52,70
Therefore E(X) = = 52,70
Var(Y) = Var(x−50) = Var(x)
Va(Y) = = = = 18,29

Example 8
Below is a sample number of children taken from a population of families in a
certain village.
5 3 2 4 1 3 5 6 2 4 5 6 8 4 7 2 1 3 5 9 Calculate the unbiased estimate
of population proportion of family with more than 5 children

SOLUTION
Sample proportion of families with more than 5 children = =
= =

COMPILED BY: MANYUVIRE D

COMMENTS ON ERRORS AND OMISSIONS
ARE WELCOME.
CALL ,TEXT OR WHATSAP 0783235483
ONLINE LESSONS ALSO OFFERED

11