methods of minimizing these errors.
Q 1: What is sampling? Why is it
important in psychological research?
Sampling
is the method of drawing an inference about the characteristics of the
population or universe by observing only a part of the population.
To
understand sampling in true terms, it is important to understand a few
fundamentals of sampling. These are:
a)
Population: A
population may be defined as any identifiable and well-specified group of
individuals to which findings of a survey research are to be extrapolated. A
population can be finite (one of which
all the members can be counted) or infinite
(one whose size is unlimited and therefore, the members cannot be counted).
Population: A
population may be defined as any identifiable and well-specified group of
individuals to which findings of a survey research are to be extrapolated. A
population can be finite (one of which
all the members can be counted) or infinite
(one whose size is unlimited and therefore, the members cannot be counted).
b)
Sample: A
sample is a part of the population being most representative of it. The sample
is drawn with the purpose of drawing a fair representative of the population
and which leads to estimates of population characteristics with great
‘precision’ and ‘accuracy’.
c)
Sampling Units and Sample frame: The
individual members of the population whose characteristics are to be measured
are called sampling units or elements
of the population. A list of all such sampling units makes up a sampling frame. In other words, a
sampling frame is a list of the elementary units from which a sample can be
drawn.
d)
Parameter and Statistic: A statistic is a numeral value which is
based upon the sample and a parameter
is a numeral value which is based upon a population. The primary purpose of any
survey research is to estimate certain values relating to the distributions of
specific characteristics of a population. These estimates are called the 
parameter. For the same population, the
parameter remains constant. However, sometimes it becomes impossible to
estimate the population parameter directly as the population size may be large
and unmanageable (e.g., all married couples on this earth). In such cases,
samples are drawn from the population and parameter is estimated by assessing
the statistics of the samples.
parameter. For the same population, the
parameter remains constant. However, sometimes it becomes impossible to
estimate the population parameter directly as the population size may be large
and unmanageable (e.g., all married couples on this earth). In such cases,
samples are drawn from the population and parameter is estimated by assessing
the statistics of the samples.
e)
Sampling design: A
sampling design is the detailed plan of obtaining a sample from the sampling
frame. It refers to the procedure the researcher would adopt for selecting a
sample from which inferences about the population are drawn.
Thus,
the sampling procedure follows defining a population, drawing a sample from the
population following a sample design and estimate the population parameter from
the sample statistics. The steps of sampling can be therefore enlisted as,
i)
Stating the objectives of the survey
ii)
Defining the population to be sampled
iii)
Determining the data to be collected
iv)
Deciding on the methods of measurement
v)
Choosing the sampling units
vi)
Selecting the sample
vii)
Organizing the field work
viii)
Summarizing and analyzing the data
ix)
Planning future survey
--------------------------------------------------------------
Psychological
research largely involves studying human beings and animals at large and
drawing inferences regarding certain behavior. However, it becomes often
impossible for the researcher to consider the whole universe for a study. For
example, suppose the researcher 
wants
to study the play behavior of 3-year old children or the drinking behavior of
adolescents or may be more specifically the emotional control of schizophrenic
caregivers. It is practically impossible for the researcher
to consider all the 3-year old children of the world or all the adolescents of
the world or even the caregivers of all people suffering from schizophrenia.
Therefore it becomes mandatory for the psychology researcher to draw
representative samples out of the populations in order to study the populations
at large. Thus sampling is very much needed in psychological research.
wants
to study the play behavior of 3-year old children or the drinking behavior of
adolescents or may be more specifically the emotional control of schizophrenic
caregivers. It is practically impossible for the researcher
to consider all the 3-year old children of the world or all the adolescents of
the world or even the caregivers of all people suffering from schizophrenia.
Therefore it becomes mandatory for the psychology researcher to draw
representative samples out of the populations in order to study the populations
at large. Thus sampling is very much needed in psychological research.
Q 2: Critically discuss different types
of sampling.
Sampling
can be categorized into two broad categories on the basis of how the sample was
selected, namely (a) Probability sampling and (b) Non-probability sampling.
(a)
Probability sampling:
Probability sampling methods are those which clearly specify the probability or
likelihood of inclusion of each element or individual in the sample. In this
method, the sample is obtained in successive draws of a unit each with a known
probability of selection assigned to each unit of the population at the first
draw. At any subsequent draw, the probability of selecting any unit from the
available units at that draw may be either proportional to the probability of
selecting it at the first draw or completely independent of it. The successive
draws may be made with or without replacing the units selected in the previous
draws.
Advantages
and disadvantages:
|
Advantages
|
Disadvantages
|
|
The
merit of this method is that the obtained samples are considered
representative, and hence the conclusions reached from such samples are worth
generalization and are comparable to similar populations to which they
belong.
|
The
demerit of the sampling method is that a certain amount of sampling error
exists because the researcher has only a limited element of the entire
population. The smaller the sample, the greater the sampling error.
|
(b) Non-probability
sampling: The non-probability sampling method is one in
which there is no way of assessing the probability of the element or group of
elements of the population being included in the sample. In other words,
non-probability sampling methods are those that provide no basis for estimating
how closely the characteristics of a sample approximate the parameters of the
population from which the sample had been obtained. This is because
non-probability samples don’t use the techniques of random sampling. In
non-probability sampling, the reliability of the resulting estimates can’t be
evaluated which results in the user not knowing how much confidence can be
placed in any interpretations of the survey findings.
Some of the probability and
non-probability sampling techniques are discussed below.
             
|
Figure 1. Types of
sampling techniques.
A.
Simple Random Sampling: The
simplest method of probability sampling is the simple random sampling, wherein,
units are drawn one by one with replacement or without replacement. Let N be the units of the population and n be the number units of the sample (n<N). There are two ways of
performing simple random sampling, viz., simple random sampling with
replacement of units and simple random sampling without replacement of units.
a) Simple random sampling with replacement: In
this method of sampling, each unit of the population has the equal probability
of being selected as an unit of the sample given by the following formula:
The probability of
selection of each unit = 
This means that the first
unit of the sample will be selected from the population with a probability of
1/N. The selection of the unit from the sample can be done using either random
number table or by using Monte-Carlo Simulation. Each of the remaining units
required for the sample is selected in the same way using the same probability
of selection. For example, if we are to select a sample of 50 out of 200
students of eighth grade of a school, we can follow any of the random selection
method to select those 50 students. As in we can write their names on slips and
shuffle and reshuffle those slips. Then we can draw out those slips one by one
and the names written on the slips will be included in the sample. So the
probability of the first slip to be included in the sample is 1/200. In the
random sampling method with replacement,
once the name written on the slip is recorded, the slip is returned in the
container and the slips are reshuffled before the next slip is drawn. So that
the probability of getting selected in the sample of the second, third, fourth
and all other slips remains 1/200.
b) Simple random sampling without
replacement: In the above example, if the slips after
recording of the name, the slips are not returned, it is simple random sampling
without replacement. In the sampling random sampling without replacement, each unit of the population has varying
probability of being selected as an unit of the sample given by the following
formula:
The probability of
selection of the first unit =
The probability of
selection of the second unit = 
The probability of selection of the nth
unit = 
The major difference between sampling with replacement and sampling without replacement is mainly
concerned with the number of possible samples of size n could be theoretically drawn. In the case of sampling with replacement the number of possible
samples of size n would be greater
than the number of possible samples of the same size (from same population) in
case of sampling without replacement.
For example, suppose the size of the population consists of 10 persons and the
researcher wants to select samples of size 5 through the procedure of sampling
without replacement. In such a situation, the researcher can maximally draw 252
samples following the formula:
= 
Where,
N =
the size of parent population
n = the
size of the sample
= factorial
In
the above example, where N = 10 and n = 5, the maximum number of sample size
of 5 would be:
= 
= 
= 
However,
if the researcher decides to proceed with the technique of simple random
sampling with replacement, he can derive the likely number of samples from the
given population with the following equation:
Nn
Following
this formula, from the same population of N
= 10 and sample size n = 5, the
researcher can therefore draw 105 = 10 × 10 × 10 × 10 × 10 = 100000
samples.
Advantages and disadvantages
of Simple Random sampling technique:
Advantages:
i)
Since this method involves random selection of
units, each and every unit of the population has an equal chance of getting
selected in the sample. The result is an unbiased
sample representative of the target population.
ii)
In such a sampling technique, the investigator
does not need to know the true composition of the population beforehand.
iii)
In simple random sampling, sampling error
associated with any given sample drawn can be easily assessed.
Disadvantages:
i)
This method does not ensure that those elements
which exist in small numbers in the population will be included in the given
sample. For example, suppose the researcher wants to study the play behavior of
children with mental retardation. So he wants to include all those children who
are mentally retarded because of different kinds of brain disorder. Now, the
prevalence of MR is among generally noted among children with cognitive
dysfunction from birth or in case of other genetic problems like Down’s
syndrome. MR can also occur in case of progressive developmental disorders like
Autism, Rett’s Syndrome, and Asparagus Syndrome etc. However, the prevalence of
these disorders especially Rett’s Syndrome is much less in the population. In
such a case, if the researcher wants to do simple random sampling, it is
unlikely that children with rare problems will be included in the sample.
ii)
In case of simple random sampling technique,
the sampling error of a sample size n is
greater as compared with the sampling error incurred in the case of other
random sampling techniques. This is because the heterogeneity in the case of a
random sample is greater than the same in the case of other random sampling
techniques like stratified random sampling. In stratified random sampling, the
sample drawn becomes somewhat typical of the population because it is more or
less proportionate to some known characteristics of the population. This fact
is ignored in simple random sampling. Hence, the sampling error increases.
--------------------------------------------------------------------
B.
Stratified Random Sampling: Stratified
sampling is an improvised sampling over simple random sampling. This sampling
will have more statistical efficiency. In this technique, the population is
divided into a specified set of strata such that the members within each
stratum have similar attributes but members between strata have dissimilar
attributes. This means that each stratum is homogenous when compared to the
population.
Suppose, a researcher wants
to study the academic achievements of eighth grade students of the rural, urban
and semi-urban areas of West Bengal. So he needs to stratify the student population
according to the areas.
 |
The objective of the stratified sampling is
to select a sample of size n from the
population such that the following condition is satisfied.
n = n1
+ n2 + n3+……nk
Where
ni is the number of
sampling units taken from the stratum i,
i = 1, 2, 3….k; k being the number of strata.
 |
The divided populations are called subpopulations,
which are non-overlapping and together constitute the whole population. Having
divided the population into two or more strata, which are considered to be
homogenous internally, a simple random sample for the desired number is taken
from each population stratum. Thus in stratified random sampling the
stratification of population is the first requirement. There can be many
reasons for stratifications in a population. Stratification tends to increase
the precision in estimating the attributes of the whole population. If the
whole population is divided into several internally homogenous units, the
chances of variations in the measurements from one unit to another are almost
nil. In such a situation a precise estimate can be made for each unit and by combining
all these estimates, we can make a still more precise estimate regarding the
population. Again, stratification gives some convenience in sampling. When the population is
divided into several units, a person or group of persons may be deputed to supervise
the sampling survey to each unit.
Stratified random sampling is of
two types:
a. Proportionate
stratified random sampling
b. Disproportionate
stratified random sampling
a. Proportionate
stratified random sampling: In a
proportionate stratified random sampling, the researcher stratifies the
population according to the known characteristics of the population and,
subsequently, random draws the individuals in a similar proportion from each
stratum of the population.
Suppose
N be the size of the population; Ni be the size of the stratum i; ni be the size of the
sub-sample to be taken from the stratum i;
k, be the number of strata in the
population and n, be the size of the
total sample of the population. Then, ni
should be decided based on the following relationship:
= 
=
=…= 
= 
And
N1 + N2 + N3 +…+
N k = N
Suppose,
total number of engineering colleges in the university, N = 200. Among these colleges, government engineering colleges, N1 = 20, number of aided
engineering colleges, N2 = 50
and number of private colleges, N3
= 130. Suppose the researcher wants to draw a sample of size n = 20. Then,
= 
= 0.10
Therefore,
n1 = 
N1
= 0.1 × 20 = 2
N1
= 0.1 × 20 = 2
n2
= N2
= 0.1 × 50 = 5
N2
= 0.1 × 50 = 5
n3
= N3
= 0.1 × 130 = 13
N3
= 0.1 × 130 = 13
Therefore,
the sample of 20 should contain 2 government colleges, 5 aided government
colleges and 13 private colleges.
Advantages
and disadvantages of proportionate stratified sampling:
|
Advantages
|
Disadvantages
|
|
1.
Increases the representativeness of the
sample drawn.
2.
Sampling error is minimized.
3.
Eliminates the necessity of weighing
the elements according to their original distribution in the population.
|
1.
It is a difficult method.
2.
It is a time-consuming method.
3.
This method has the probability of
classification error.
|
b. Disproportionate
stratified random sampling: In disproportionate
stratified random sampling, the substrata of the drawn sample are not
necessarily distributed according to their proportionate weight in the
population from which they were randomly selected. In fact, some of the strata
of the population may be over-represented or some under-represented.
Suppose,
the investigator divides a given population of 10,000 individuals into 6,000
males and 4,000 females. If he has to draw a sample of 1,000 individuals from
the set of 10,000, he can draw randomly both the males and females in equal
number, say, 500 each, it will constitute the example of disproportionate
stratified random sampling. In this example, when he randomly draws 500 males
and 500 females, he is over-representing a female stratum and
under-representing a male stratum.
Advantages
and disadvantages of disproportionate stratified sampling:
|
Advantages
|
Disadvantages
|
|
1. It
is comparatively less time-consuming method.
2. Here,
the investigator is able to give weight to the particular group of elements
that are not represented as frequently in the population as compared with
other elements.
|
1.
In this method, the population is
over-represented and some other strata are under-represented.
2.
Where, the composition of the population
is unknown to the investigator, this method cannot be applied.
3.
There remains a probability of
misclassification in this method.
|
-----------------------------------------------------------------
C. Systematic
sampling: This is a special kind of random sampling in
which the selection of the first unit of the sample from the population is
based on randomization. The remaining units of the sample are selected from the
population at a fixed interval of n, where
n is the sample size.
Let
the size of the population (N) be 800
and the sample size (n) be 40. The
units of the sampling frame are divided into n number of intervals based on the ratio N/n, as shown below:
Sampling
interval width, I = 
= 800/40 = 20
= 800/40 = 20
The
sampling frame consists of units with serial numbers from 1 to 800. This range
is divided into 40 intervals, viz., 1-20, 21-40, 41-60, 61-80,…, 760-780 and
780-800, where the total number of intervals is equal to the sample size.
Then
a number from the first interval 1-20 is selected randomly and the unit of the
population with this serial number is treated as the first unit of the sample.
Let the randomly selected unit the first interval of the population be 12.
Then, the second unit of the sample is the unit in the population with serial number
32 which is obtained by adding 20 (sampling interval width, I) to 12. Then, each of the remaining
units of the sample can be obtained from the population in the same manner by
adding 20 to the serial number of the previous unit selected from the population. As per these guidelines, the units of the
population with serial numbers 52, 72, 92…772 and 792 are treated as the third,
fourth, fifth,…, 39th and 40th units of the sample
respectively. Sometimes, it may happen
that I may not be an integer (whole
number), say 19.25. in that it becomes cumbersome to use the systematic
sampling method. A solution to this problem is the use of the Circular
Sampling method. In this method, instead of considering the sampling
frame to be a linear list, it is considered as a circular list such the last
unit is followed by the first unit. Also, in this method, the initial point is
any number between 1 to N. Thus, in
the above example, any number between 1 and 800 is chosen, say 256 and then the
second number becomes 256 + I =276,
the third number becomes 276 + I =
296 and so on.
Advantages
and disadvantages of systematic sampling:
|
Advantages
|
Disadvantages
|
|
1.
Relatively quick method.
2.
Facilitates easy counting of the sample
units included.
3.
Easy to use.
|
1.
Ignores all units which are in between
every nth element chosen.
2.
Sampling error may increase.
|
-------------------------------------------------------------------
D. Cluster
Sampling: Cluster
sampling
is a sampling technique in which the population is divided into clusters such
that the members within each cluster are dissimilar (heterogeneous) in terms of
their attributes, but different clusters are similar to each other. This leads
to the inference that each cluster can be treated as a small proportion which
possesses all the attributes of the population. Hence, in cluster sampling, any
one of the clusters is randomly selected and all the units of that cluster are
selected (sampled) to arrive at inference about the population.
Suppose,
the researcher wants to study the culture of the population of a state and its
impact on the economy of the state. For this, the researcher needs to sample
from the different geographical regions of the state as in:
 |
Here, one can conveniently assume that these
clusters (1 to 10) are similar to each other and the members within each
cluster are heterogeneous. This kind of sampling can be called area sampling as the population and
clusters are defined with reference to the geographic regions.
Advantages
and disadvantages of cluster sampling:
|
Advantages
|
Disadvantages
|
|
1. It
is easier to use when large geographical areas are to be covered.
2. Respondents
can readily be substituted in the same random section.
3. Economical
in time and money.
4. Flexible
in nature.
|
1.
Degree of sampling error is usually
high.
2.
Here the researcher has little control
over the size of each cluster.
3.
It is difficult to ensure that the
individuals included in one cluster are independent of other randomly drawn
clusters.
|
--------------------------------------------------------------
E. Multistage
sampling: In a large scale survey covering the entire
nature/subcontinent, the size of the sampling frame will be too large which
leads to more time and cost of the study. In such study, multistage sampling
technique helps designing a smaller frame which will make the study practicable
in terms of cost and time.
The
multistage sampling employs more than one stage to sample the population
depending upon the reality. The combination of the types of sampling techniques
to be used in the specified number of stages is unique to the reality.
For
example, suppose a researcher wants to study the infrastructure of schools in a
country. For this study, one can use the multistage sampling technique. In
stage I, the researcher may stratify the whole country into different regions
as in east, west, north and south and then select states from these regions
(strata). Here it is assumed that the states (sampling units) within each
region are similar and the regions are dissimilar. In the stage II, the
researcher can do cluster sampling to identify a district from each selected
state assuming different districts of each state as its clusters. Here, it is
assumed that the districts of a state are similar, but the schools in each
district are dissimilar in terms of their present infrastructure. In stage III,
from each selected district, a random sampling may be used to select
proportionate number of schools (sampling units) from it. Therefore, the study
uses three stages of sampling. The highest levels of sampling unit are states
and the lowest levels of sampling unit are schools. Thus, multistage sampling
reduces the size of the overall sampling frame.
------------------------------------------------------------------------
F. Convenience
sampling: Also called accidental sampling or incidental
sampling, this is a non-probability method in which the investigator
selects the sampling units based on his convenience. Here, the investigator
does not care about including units with some specific or designated trait;
rather he is mainly guided by convenience and economy.
For
example, the researcher may take the students of class X as because the class
teacher of that class is his friend. In psychological research, often we use
this method.
Advantages
and disadvantages of convenience sampling:
|
Advantages
|
Disadvantages
|
|
1. Saves
time, money and labor of the researcher.
|
1. Generalization
cannot be done with confidence.
2. May
get affected by researcher’s bias and prejudice.
3. Sampling
error is high.
|
----------------------------------------------------------------
G. Purposive
or Judgment sampling: This is also a non-probability sampling
method in which the sampling units are selected on the advice of some expert or
by the intuition/opinion of the researcher himself. In the first case, an
expert who is familiar with the sampling frame guides the researcher in
selecting the sampling units from the sampling frame. For example, suppose a
researcher wants to study the reasoning ability of the students of a particular
state. He might go to the school education council of the state government and
seek the advice of the director for selecting schools. In the second case, the
researcher applies his/her intuitive judgment and previous experience in
selecting the sampling units from the sampling frame. For example, suppose a
researcher wants to study the attitude towards political issues of a country.
For this study, he might select journalists, legislators and teachers because
they can reasonably be expected to represent the correct attitude.
Advantages
and disadvantages of purposive sampling:
|
Advantages
|
Disadvantages
|
|
1. Less
costly and more readily accessible.
2. Guarantees
that only those individuals who are relevant for the study will be included
and not others.
|
1. There
is no way to ensure the representativeness of the sample.
2. This
method is quite subjective in nature.
|
-----------------------------------------------------------------------
H. Quota
sampling: In quota sampling (another non-probability
method), the investigator recognizes different strata of population and from
each stratum he selects the number of individuals arbitrarily. It is similar to
proportionate stratified random sampling in that both the methods follow the
selection of sampling units from strata in equal proportion as they exist in
the population. But they are different in that, quota sampling does not follow
the rule of random selection of units.
For
example, let the percentage of old people in the population be 20% and that of
middle age and young age be 50% and 30% respectively. In the quota sampling,
the proportions of the number of sampling units selected from these categories
are same as they exist in the population. If the researcher wants to study the
health behavior of such people, he will have to draw sampling units from these
categories in equal proportion as they exist in the population. So the
proportions would be 0.2n, 0.5n and 0.3n respectively.
Advantages
and disadvantages of quota sampling:
|
Advantages
|
Disadvantages
|
|
1. It
is a satisfactory mean when quick and crude results are desired.
2. Guarantees
the inclusion of individuals from different strata of the population.
|
1. Lacks
external validity.
2. Sample
selected is not very representative of the population.
3. Classification
error may occur.
4. Is
a less dependable method.
|
-----------------------------------------------------------------
I. Snowball
sampling: The snowball sampling is a restrictive multi
stage sampling in which initially a number of sampling units are randomly
selected. Later, additional sampling units are selected based on referral
process. This means that the initially selected respondents provide addresses
of additional respondents for the investigator.
For
example, suppose a researcher wants to study the personality profile of
terrorists. For this purpose, the researcher might at first identify one
potential respondent. Further, he might obtain the contacts of other terrorists
residing in the area from that potential respondent and the process continues
until the researcher obtains data from desired number of respondents. The
snowball technique is very much useful for studying such hidden populations.
Advantages
and disadvantages of snowball sampling:
|
Advantages
|
Disadvantages
|
|
1. Very
much useful in studying small informal social groups.
2. Reveals
communication patterns and decision making techniques used in community
groups.
|
1. Becomes
cumbersome and difficult when N is
large.
2. May
be biased and not the true representative of the population.
|
--------------------------------------------------------------------------
J. Saturation
sampling and Dense sampling: These two non-probability methods were
emphasized by Coleman (1959). Saturation
sampling is defined as drawing all elements or individuals having
characteristics of interest to the researcher. For example, drawing all the
teachers having at least 10 years of teaching experience (from a particular
school zone) for a study may be called saturation sampling. Dense sampling is a method of sampling
which lies somewhere between simple random sampling and saturation sampling.
Here, the researcher may select 50% or more from the population and takes a
majority of individuals having a specified trait or characteristic which are of
interest to him. For example, the researcher studying the word salad problem in
schizophrenia may select 50 to 60 such patients with the symptom of word salad
from among a population of 100 schizophrenia patients. Both the saturation
sampling method and dense sampling method becomes cumbersome when the size of
the population is very large.
-------------------------------------------------------------------------------
K. Double
sampling: In sampling, the use of a ratio or regression
n estimate with an independent variate, highly correlated with the variate
being estimated, can sometimes considerably increase the precision of results.
Similarly, the precision is increased by stratification on one or more variate
if there is a high correlation between the stratification groups and the
variate being estimated. In order to be used in that way, the independent
information must be available only from a sample; if not readily available, it
can sometimes be collected for a comparatively large sample at a very low cost.
When this is so, Neyman (1938) has shown that it may pay to select an initial
large sample in order to obtain the low cost information, and to draw a sub
sample from the larger sample on which measurements are made of the desired
characteristics. The information from the large sample may be used either in
ratio, difference or regression estimates or for stratification, in order to
increase the reliability of the desired estimates from the smaller and more
costly sample. This is the double
sampling method. It will pay to use such a double sampling method only if
the cost of the initial sample is small and the gains from regression
estimation or stratification are large.
For example, suppose
the researcher wishes to study the reasoning ability of patients suffering from
obsessive compulsive disorder. For this, he collects data from an initial
sample of say 500 patients. He finds that the reasoning ability differs among
the male and female patients and also among the patients with the awareness of
obsession and without awareness of obsession. Based on these information, the
researcher further stratifies the sample and selects say 100 patients, 50 male
(25 patients with awareness and 25 patients without awareness) and 50 female
(25 patients with awareness and 25 patients without awareness) patients for
further data collection.
Q 3:
What is sampling error? How do you minimize it?
Sampling
Error:
In sampling, whatever is the method of
selection, a sample estimate will inevitably differ from the one that would be
obtained from enumerating the complete population with equal care. This
difference between the sample estimate and the population value is called the sampling
error. The larger the sample the smaller will be the sampling error on
the average and greater will be the confidence in the results. Heiman (2002)
defined sampling error as "the difference, due to random chance, between a
sample statistic and the population parameter it represents".
Sampling
error can be of two types, namely, random
error and systematic error.
i.
Random error is a
pattern of errors that tend to cancel one another out so that the overall
result still accurately reflects the true value. Every sample design will
generate a certain amount of random error.
ii.
Systematic error or Bias, on the other hand, is more serious because the pattern of
errors is loaded in one direction or another and therefore do not balance each
other out, producing a true distortion.
-------------------------------------------------------------------------------------
Methods
to minimize sampling error:
There
are various rules by which one can reduce the sampling error. These are:
a.
Using considerably large sample size. As
the size increases, the sample gets closer to the actual population, thereby
decreasing the potential for deviations from the actual population.
b.
Another potential method of minimizing the
sampling error the selection of the sample through probability sampling. Here,
every unit of the population has an equal chance of getting selected in the
sample, thereby reducing the bias in selection procedure.
c.
Stratification is another
method of obtaining greater precision in our sample estimates. In this method,
secondary information can be utilized to divide the population into groups such
that the elements within the each group are more alike than are the elements in
the population as a whole. This ensures greater precision in the estimation of
population parameter from the sample statistics.
Stratification is another
method of obtaining greater precision in our sample estimates. In this method,
secondary information can be utilized to divide the population into groups such
that the elements within the each group are more alike than are the elements in
the population as a whole. This ensures greater precision in the estimation of
population parameter from the sample statistics.
Bibliography
and References:
·
Cochran, W. G. (1953). Sampling Techniques. New York: John Wiley & Sons.
·
Coleman, J. S. (1959). Rational analysis: The
study of social organization with survey methods. Human Organization, 17: 28-36.
·
Hansen, M. H., Hurwitz, W. N. & Madow, W.
G. (1993). Sample Survey Methods and
Theory, Vol. 1: Methods and Applications. New York: John Wiley & Sons.
·
Heiman, G. W. (2002). Research Methods in
Psychology. 3rd Edition. Boston & New York: Houghton Mifflin Company.
·
Neyman, J. (1938). Contributions to the Theory
of Sampling Human Populations. Journal of
American Statistical Association, 33, pp: 101-116.
·
Panneerselvam, R. (2013). Research Methodology. Delhi: PHI Learning Private Limited.
·
Singh, A. K. (2011). Tests, Measurements and Research Methods in Behavioral Sciences;
Bharati Bhawan.
·
Sukhatme, P. V. (1953). Sampling Theory of Surveys with Applications. Iowa, USA: The Iowa
State College Press.
·
Yates, F. (1949). Sampling Methods for Censuses and Surveys. London: Charles Griffin
& Company.
No comments:
Post a Comment