Sampling: Simple Random, Convenience, systematic, cluster, stratified - Statistics Help

TEACHER: Sampling-- Simple,
Convenience, Systematic,

Cluster, and Stratified.

To find things out about
a population of interest,

it is common practice
to take a sample.

A sample is a selection of
objects or observations taken

from the population of interest.

For example, a population
might be all apples

in an orchard at a given time.

We wish to know how
big the apples are.

We can't measure all of them.

So we take a sample of some
of them and measure them.

The method chosen
for taking the sample

depends on the nature
of the population

and the resources available
in terms of time and money.

The ideal is for each
object in the population

to be equally likely to be
chosen as part of the sample.

This is called an
unbiased sample.

It is also desirable
for the sample

to be representative
of the population.

If the population of apples
were 2/3 red and 1/3 green,

the sample should
be similarly split.

Note that no matter what
we do, there will always

be sampling error or
variation due to sampling,

as we are looking at a
part of the population, not

the whole population.

The video on variation covers
these concepts more thoroughly.

This video presents five
methods of sampling--

simple random sampling,
convenience sampling,

systematic sampling, cluster
sampling, and stratified

sampling.

For each method, we will outline
the process and the advantages

and disadvantages.

Simple random sample--
simple random sampling

is theoretically the
ideal method of sampling.

You list each member
of the population

and use random numbers
to decide which

objects are in the sample.

Each object is equally
likely to be selected.

This produces an
unbiased sample, which

we hope is representative.

However, it can be difficult
and expensive to take

a simple random sample
when dealing with people.

Simple random sampling
is more practical

when a population is
geographically concentrated

and when a good
sampling frame exists.

A sampling frame is a list
of all the people or objects

in the population of interest.

Simple random sampling can
be more easily implemented

for natural and
manufacturing populations.

Convenience sample--
convenience sampling

is just that, convenient.

You ask people
nearby or people who

walk past at a shopping mall.

Or you take the next 20 objects
off the production line.

You do what is
easy or convenient.

Convenience samples are
often biased in some way.

But for a quick and cheap
poll, it may not really matter.

Convenience samples can also
have self-selection bias

when people choose
to participate

because they have an interest
in the issue in question.

Systematic sampling--
with systematic sampling,

you choose a starting
point at random

and then systematically
take objects

at a certain number apart.

For example, if there are
1,000 in the population

and you want a sample of 50, you
would take every 20th object.

Systematic samples are
easier to administer

than simple random
samples and are

usually a good approximation
of a random sample.

However, if there is a
pattern in the population,

certain types of objects could
be chosen more or less often

than others.

Cluster sampling--
in cluster sampling,

the population is divided
into clusters, which

are then chosen at random.

For example, departments
of a business

can be clusters or
suburbs within a city.

Within each cluster,
all of the objects

are included in the sample.

Cluster sampling can be more
convenient and practical

than simple random sampling.

However, if the clusters are
different from each other

with regard to the
elements we are measuring,

it can lead to bias or
non-representativeness.

Stratified sampling--
stratified sampling

seems like cluster sampling.

But the strata, or groups,
are chosen specifically

to represent different
characteristics

within the population, such
as ethnicity, location, age,

or occupation.

Within each group,
a random sample

is taken, sometimes
in proportion

to the size of the group.

Stratified sampling
can lead to a very good

random representative sample.

However, it can be
complex to administer.

And a sampling frame with
considerable information

about the population
is required.

There are other
sampling methods.

The five explained
here give an idea

of the advantages
and disadvantages

of various methods.

You should attempt to use
the sampling method that

produces the best result for the
resources you have available.

If your sample has
known bias, this

should be taken into account
in analysis and reporting.