Professional Documents
Culture Documents
MARKER’SCOMMENTS_____________________________________________________
___________________________________________________________________________
________________________________________________________________
Introduction
Sampling is the method of selecting a representative subset of the population called sample.
Sampling makes research more accurate and economical. It’s the sampling method which
actually determines the generalizability of the research findings. In simple words, the process
of choosing a sample of the population to study is called sampling. For example, you wish to
study newspaper reading habits among the students of at a college. It is not possible to
include all the students in your study. Therefore, it is necessary to draw a sample of the whole
population - all the students of the university. Suppose, from a total of 500 students at a
college, select a sample of 50 students to trace out whether they read a newspaper or not? To
calculate the portion of students qualified to be the part of the sample, there is need to divide
sample size by population size; 50/500 = 0.1 %. This reveals 10 per cent of the population is
to be studied.
Probability Sampling
In probability sampling, each sample has an equal probability of being chosen. It is the
probability sample is one in which each element of the population has a known non-zero
probability of selection. This method of sampling gives the probability that the sample is
representative of a population. For example 500 students at a school have an equal chance to
be the participants in a study. Probability sampling uses random sampling techniques. For
example, if there was a population of 100 people, each person would have odds of 1 out of
100 of being selected. In this sampling method, there is an utmost probability of a
representative sample of the whole population.
Cluster Sampling
Cluster sampling is one of the efficient methods of random sampling in which the
population is first divided into clusters, and then a sample is selected from the clusters
randomly. Unlike the above, in pure cluster sampling, the whole cluster is sampled. In
contrary to stratified sampling, there should be heterogeneity within the clusters and
homogeneity between the clusters. The more homogeneity among the clusters, lesser will be
the margin of error or vice-versa. The method is mostly feasible in case of diverse population
spread over different areas. Not possible in case of simple random sampling, the
participants with different demographics are selected randomly from these areas. This is
an economical method and saves the time of researcher. Suppose, a researcher wants to study
the average woman’s height in Zimbabwe. It is not possible to get a list of everyone. But in
such a case, the researcher will divide the areas into clusters and draw a sample
randomly from those clusters.
Multi-Stage Sampling
To draw the sample, this method actually uses a combination of various techniques. In this
method, the population is divided into groups at various levels that is a group within a group,
within a group and so on. The sample is finally drawn from the smallest group among all the
groups.
Non-probability Sampling
Unlike probability sampling method, non-probability sampling technique uses non-
randomised methods to draw the sample. Non-probability sampling method mostly involves
judgment. Instead of randomisation, participants are selected because they are easy to access.
For example; a researcher’s classmates and friends have a better chance to be part of his or
her sample. Even though in certain cases, non-probability sampling is a useful and convenient
method of selecting a sample, the method is appropriate and the only method available in
certain cases.
One of the major shortcomings of the non-probability sampling is that the findings
established through this method lack generalizability. Even though findings obtained through
this method apply mostly to the group studied, it may be wrong to extend these findings
beyond that particular sample. The non-probability sample is used to study existing
theoretical insights or developing new ones. This method of sampling is considered less
expensive, less complicated and easy to apply as compared to its counterpart.
Convenience Sampling
In this type of sampling, researchers prefer participants as per their own convenience. The
researcher selects the closest live persons as respondents. In convenience sampling, subjects
who are readily accessible or available to the researcher are selected. For example the
researcher may select classmates and friends for the study. In other words, in this type of non-
probability sampling method, whoever meets the researcher qualifies to be the part of the
sample. For example; people in the streets. To get the questionnaire filled, the researcher
takes copies of the questionnaire and distribute from the street and those who will be passing
by.
Purposive Sampling
In this type of sampling, the researcher chooses the participants as per his/her own judgment,
keeping back in mind the purpose of the study. It requires the judgment of an expert in
selecting cases or also selects cases with a specific purpose in mind. This type of sampling is
used in exploratory research or in field research. With purposive sampling, the researcher
hardly knows whether the cases selected do represent the population or not. As mostly
sampling problem is being addressed with a specific plan in mind, most of the sampling
methods may be considered purposive in nature. Purposive sampling is somewhat less
costly, more readily accessible, more convenient and select only those individual that are
relevant to research design. Besides this, there is no way to ensure that the sample is truly
representative of the population, and more emphasis is placed on the ability of the researcher
to assess the elements of the population. For example: For studying attitude toward any
national issue, a sample of journalists, teachers and legislators may be selected for the
study. They qualify to be part of purposive sampling as they can more reasonably be
expected to represent the correct attitude than other class of people.
Quota Sampling
This is considered as being the non-probability equivalent of stratified sampling. The purpose
here is to draw a sample that has the same proportions of characteristics as the whole
population. The procedure relies on accidental choice instead of relying on random choice.
Although much less accurate than stratified sampling - since it is not based on a random
sample - the method is often more convenient and economical that is no lists have to be
compiled and all data can be collected at an arbitrary location.
Snowball Sampling
Also called "chain referral sampling,” in this method, the sample is actually collected in
various stages. Snowball sampling which is a non-probability sampling method is basically
socio metric in nature. Although snowball sampling is considered to be a form of accidental
sampling by some, this method is appropriate when the members of a special population
are difficult to locate for example homeless people, migrant workers etc. It begins by the
collection of data from one or more contacts usually known to the person collecting the data.
At the end of the data collection process (e.g questionnaire, survey, or interview), the data
collector asks the respondent to provide contact information for other potential
respondents. These potential respondents are contacted, interviewed and further asked to
provide more contacts. This process goes on till the purpose of the researcher is achieved.
According to Daniel (2012), “In a sense quota sampling combines availability sampling
and purposive sampling by targeting specific numbers of elements that have specific
characteristics.” Snowball sampling is most useful when there are very few methods to secure
a list of the population or when the population is unknowable.
2. A researcher is looking at purchasing behaviour of TVs over
(a) The last 2 years and an acceptable sampling error level is set of ± 5 % ( 0.05 ) . The
researcher is confident at 95% confidence level. He estimates that the estimated
percentage of the population that has bought TVs over the last 2 years is 25%.
Calculates a random sample size. [7]
Solution
Sample Size = ( ez )
2
= ( 1,96
0,05 )
2
= 39,22
= 1 537
(b) In a grocery purchases study, the standard deviation was estimated as being
$40.50. The researcher decides that she needs to be 95% confident that a true
population mean falls into the interval defined by the sample mean plus or minus
$5. Calculate a random sample size. [7]
Solution
Sample Size = ( z ×e σ ) 2
= ( 1,96 ×40,50
5 ) 2
= 15,8762
= 252
(c) Discuss 5 factors that can affect the determination of a sample size. [11]
Introduction
Sample size is one element of research design that researchers need to consider as they plan
their study. Reasons to accurately calculate the required sample size include achieving both a
clinically and statistically significant result and ensuring research resources are used
efficiently and ethically. Study participants consent to study involvement on the basis that it
has the potential to lead to increased knowledge of the concept being studied, however if a
study does not include sufficient sample size to answer the question being studied in a valid
manner, then enrolling participants may be unethical. Although sample size is a consideration
in qualitative research, the principles that guide the determination of sufficient sample size
are different to those that are considered in quantitative research.
Effect size
This is the difference or change expected in your study primary outcome as a result of the
intervention being delivered. In order to determine effect size it is essential that the primary
outcome being measured is clearly defined. A primary outcome can be collected and
measured in a variety of ways, with some examples including physiological data such as
blood pressure or heart rate, instrument scores such as quality of life scores or time to event
data such as length of stay or survival time.
The sample size calculation should be based on the primary outcome measurement. After a
relevant primary outcome measurement has been identified, the expected difference or effect
size in that outcome is estimated. Determining an expected difference can be achieved by
examining pre-existing data, for example from a previous study or pilot studies or from
routinely collected data such as quality audit data. In general, the smaller the anticipated
effect size is (i.e. the smaller the difference between groups), the larger the required sample
size. For example, if the primary outcome is incidence of delirium and pilot data suggests the
intervention is likely to reduce the incidence from 80% to 40%, this will require a smaller
sample size than an outcome such as incidence of central line infections where an
intervention might be expected to reduce the rate of infection from 5% to 4%.
The risk of error
There are two aspects to consider including the level of significance and the power. The level
of significance (referred to as α) defines the strength of identifying an effect when no effect
exists, in other words having a false-positive result. A type I error (false-positive) occurs
when we wrongly conclude there is a difference, i.e. with an α of 0.05 there is a 5% risk of a
false-positive result. The lower the level of α the less likely it is that a type I error will occur.
When determining the appropriate level of significance it will be necessary to consider the
potential impact of a false-positive result; if the potential impact is serious then a lower level
of significance, for example, α= 0.01(1% risk), might be selected. 5
This determines the likelihood of not detecting an effect when an effect does actually exist in
hypothesis testing studies, in other words having a false-negative result (type II error). A type
II error (or false-negative) occurs when we wrongly conclude there is no difference. The
higher the power of a study the less likely it is that investigators will fail to detect an effect
when an effect does exist. The power of the study is equal to 1-β, where β is the level of
acceptability of a false negative result (β =0.20 or 20% is typical, giving a power of 80%),
however in a similar fashion to level of significance it may be increased or decreased based
on the potential impact of this type of error.
The researcher is able to determine the sample size due to the type of study being conducted.
For example sample size for the average height of women in Zimbabwe will differ from the
sample size of the average height of women in Marondera. Both studies require samples from
different populations therefore method of research and sample size will differ. Also a
research at a particular organisation may cause the sample size to differ from another
organisation.
References
http://optimierung.mathematik.unikl.de/mamaeusch/veroeffentlichungen/ver_texte/sampling
_en.pdf.
http://www.statisticshowto.com/probability-sampling/