Professional Documents
Culture Documents
What is Element?
An element is the object about which or from which the information is desired.
Examples of Element:
Each Rickshaw puller of J.U
Each under five children of J.U campus.
Each sex worker of a brothel
Planning a Survey
Target population
The frame
Sample design
Method of measurement
Selection and training
The protest
Organization of field work
Organization of data management
Data analysis
Method of Sampling
Sampling techniques may be broadly classified as
1) Non Probability sampling
2) Probability sampling
1) Non-Probability Sampling:
Non-probability sampling relies on the personal judgment of researcher. Rather than chance to select
sample elements. The researcher can consciously decide what elements to include in the sample. Non
probability samples may yield good estimates of the population characteristics. But they do not allow
for observation lie evaluation of the precision of the sample results and to test-statistically.
a) Convenience Sampling:
Non-probability samples that are restricted are known as convenience samples. Restricted or field
workers have the freedom to choose whomever they find; thus the name convenience. Undoubtedly, it
is simplest and least reliable form of non probability sampling. The primary virtue is its low cost.
b) Judgment Sampling:
Judgment sampling or expert choice is one in which cases are included for investigation through a
planned selection procedure.
In judgment sampling, individuals are selected who are considered to be most representative of the
population as a whole. It is a judgment sampling because choice of the individual units depends on
entirely on the sampler.
c) Quota Sampling:
Quota sampling is a non-probability sampling, equivalent to a stratified sampling, in which the
interviewers are told to contact and interview a certain number of individuals from certain sub-groups
or strata of the population to make up the total sample.
For example, if it is known that one-third of the population lives in urban areas and two-thirds in rural
areas, the sample can be selected purposively from urban and rural areas in the same proportion. Thus a
total of 300 respondents would mean 100 urban residents and 200 rural residents to be included in the
study.
d) Snowball Sampling:
Snowball sampling is a non-probability sampling in which persons initially chosen for the sample are
used as informer to locate other persons having necessary characteristics making them eligible for the
sample through referral network.
Snowball sampling has been particularly used to study drug cultures, heroine addiction, teenage gang
activities etc.
2) Probability Sampling:
In probability sampling units are selected by chance or probability because sample elements are
selected by chance. It is possible to determine the precision of the sample estimates of the characteristic
of interest. Confidence intervals, which contain the true population value with a given level of certainty,
can be calculated. This permits the researcher to make inferences or projections about the target
population from which the sample was drawn.
Probability sampling techniques (methods) very in terms of sampling efficiency. Sampling efficiency is
a concept that reflects a trade off between sampling cost precision. Precision refers to the level of
uncertainty and the characteristic that being measured. Precision is inversely related to sampling errors
but positively related to cost. The greater the precision the greater the cost.
Non-sampling Error:
Non-sampling error can be attributed to sources other than sampling and they may be random or non-
random.
Non-response Error:
Non-response error arises when some of the respondent included in the sample does not respond.
Response Error
Response error arises when respondent give inaccurate answer or their answers are misreported.
Measurement Error:
By measurement error we understand & determining the true value of a variable or category of an
attribute of interest. If we do not do that we face measurement error.
are assumed to defer from the true values in the sample such that
where is the difference between and i.e., . The estimator of the mean per
element would then be rather than . If , then the error of may be written
where the first term on the right is the sampling error and the second the measurement error.
i) Sampling Errors:
The error, which arises entirely due to sampling and no other reasons can be attributed to cause such
error, is called sampling error.
In other words, sampling errors arise due to the fact that only a part of the population (i.e., sample) has
been used to estimate population parameters and draw inferences about the population. As such the
sampling errors are absent in a complete enumeration survey.
Total
Mean
The factor in by which the sample total is multiplied is sometimes called the expansion or
raising factor. Its inverse is known as sampling fraction and is usually denoted by i.e., .
In fact, it appears more convenient in survey sampling to use a modified definition of the population
variance rather than . The modification is trivial provided the population size is large, for it
consists of replacing by :
Theorem:
The sample mean for a simple random sample of size is an unbiased estimator of the population
mean . Symbolically,
Proof:
By definition
Taking expectation
Now,
We now need to evaluate .By Property of simple random sampling we know the probability of
Hence
Substituting this in
Corollary:
is an unbiased estimator of the population total. Since , it follows that
And
Where
Proof:
By definition
The first term of will consist of variance terms, and for each of these terms, the expected value
is the variance of the population elements. Thus,
Now consider the second term in the right-hand member of . When the sampling is without
replacement, the probability of obtaining on the draw is and the probability of obtaining
on the draw, knowing that has drawn is . Hence the probability of obtaining
Therefore
Thus
Then, because
and consequently
The variance of above based on samples without replacement differs from those with replacement
and is less than for any such that . Therefore, variance of without replacement is
less than the variance of with replacement. That is
In other words, for the same sample size, the simple estimator tends to vary less around the population
characteristics under sampling without than under with replacement. In this sense, then, sampling
without replacement should be preferred over sampling with replacement.
The factor is a correlation factor for the finite size of the population and is called Finite
Population Correlation . The sampling fraction is small when either the sample is small
or the population is large. In either case, the factor approaches and can be ignored. In such
cases, the variance of does not depend on , and there is little or no practical difference between
the two methods.
Theorem:
In simple random sampling of elements without replacement, is an unbiased estimate of .
Expressed symbolically
Proof:
By definition
Corollary:
In sampling with replacement,
That is,
So,
where
Proof:
Let , so that . The corresponding population mean of is .
Since
and
Now the second term of the right hand member of can be expanded as
Equating and
On simplification
Now by definition
Since in simple random sampling, the correlation between the means and is independent of the
sample size and is equal to the correlation between the individual observations.
Theorem:
From a simple random sample of size drown from units, simple random sub-sample units is
duplicated and added to the original sample. Show that the mean based on units is an unbiased
estimate of the population mean.
Proof:
Let represents values of the variable in the units of the population and let
from values . Without any loss of generality we can assume that the sub sample
The estimate of the population mean based on the sample of size is given by
Advantages:
Simple random sample is simple to conceptualize
Simple random sample provides foundation for much of statistical theory
Simple random sample provides a basis to which other methods can be compared
Disadvantages:
In simple random sample all enumeration units in the population must be identified and
labeled prior to sampling. This process is potentially so expensive and time consuming that it
becomes unrealistic to implement.
Sampled individuals may be so widely dispersed that visiting each selected individual may be
extremely expensive and time consuming
Certain sub-groups in the population may be totally overlooked or may be over-represented in
the sample as a result of chance factor. In either case, the estimated parameters are likely to be
in error.
When the population measurements vary considerably in size, then simple random sample
produces larger variances than other methods of sampling.
and
Then
and
From and ,
If is replaced by
when is large,
Theorem:
The variance of the sample proportion is
Proof:
Remark:
Corollary:
If , the variance of is
Proof:
Since , .
and
On the based of our answer to these five questions we can calculate the sample size needed to measure
a given proposition with a given degree of accuracy at a given level statistical significance by using a
simple formula provided that population is and above.
If the population is less than , then should be adjusted by finite population correction factor
And becomes
where is the square of the coefficient of variation of the variable. In particular, when
then for
This formula gives the sample size in for estimating with confidence coefficient
and margin of error , provided is large.
In general, if is small then the statistic if defined in equation follows student’s distribution
with . If is the significant value of distribution for and level of
significance then is given by the equation,