Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Question 3 and 4
Question 3:
Differentiate between Restricted and unrestricted sampling.
Sampling is the process of selection of a subset of individuals from within a statistical
population to estimate a characteristic of the whole population.
RESTRICTED AND UNRESTRICTED RANDOM SAMPLING
Unrestricted sampling occurs when elements are selected individually and directly from the
population, whereas, restricted sampling occurs when elements are chosen using a specific
methodology as in probability sampling or complex probability sampling.
Restricted Random Sampling:
A method of sampling is described which is a compromise between systematic sampling and
stratified random sampling. It has less potential for bias than systematic sampling and avoids
the practical problems associated with stratified random sampling.
Methods of Restricted Random Sampling: There are three methods of restricted random
sampling.
(i) Stratified Sampling:
In this method, the total population is divided into different groups or classes, which are called
Strata. A sample is drawn from each stratum. The advantage is that an efficient sample is
obtained by this method. However, note that there must be great homogeneity within each
stratum. Also, note that there should be clear-cut differences between the strata.
The method of stratified sampling has many advantages. Firstly, it is more representative of the
entire population. That is because we are dividing the population into homogeneous groups so
that none of the groups may be missed. Secondly, calculation accuracy is maximum, if the strata
are homogeneous. Thirdly, different strata can be selected from one geographical area only;
that would mean savings in terms of cost and time.
But this method is not sans limitations either. Firstly, if the strata were not homogeneous, the
results obtained from the analysis of the sample would not be reliable. Secondly, it
may not be possible to select items from each stratum on a random basis because of the
absence of skilled researchers.
In the process of systematic sampling, one unit is selected at random from the population.
Additional units are selected at evenly spaced intervals until the sample has been formed. For
this purpose, a complete list of the population should be made available to the researcher.
This list may be prepared in an order-alphabetical, geographical, numerical, or some other. All
the items are serially numbered. The first item is selected at random by the lottery method.
Then, subsequent items are selected by taking every Kth item from the list where K is K = n/N.
Here, K is called Sampling Ratio, N is the size of the universe and n is the size of the sample.
There are many advantages of systematic sampling. Firstly, it is a simple and convenient
method. Secondly, the time consumed by the researcher in sampling procedures is less.
Thirdly, results obtained after analyzing the samples (obtained by this method) are satisfactory
if care is taken to ensure that there are no periodic features associated with the sampling
interval).
Fourthly, if populations under study are sufficiently large, then this method of sampling can
give results like those obtained through a stratified sampling procedure. However, the
prominent limitation of this method in that it is less representative of the population if our
population has periodicity hidden in it, or if the sample is taken by following a periodicity trend.
In this method, random selection is made of primary, immediate and final units from a given
population or stratum. There are several stages in this sampling method.
To begin with, the first-stage units are sampled by the random sampling method. Then, a
sample of second-stage units is selected from the first stage units. The method of selection of
second-stage units may be like or different from the method used for first stage units.
We can add more stages if needed.
There are many advantages of multistage sampling. Firstly, it is a flexible method; flexibility is
lacking in other methods of sampling. Secondly, this method enables the researcher to use
existing divisions and subdivisions of the population at different stages.
Thirdly, fieldwork can be concentrated, and large areas can be covered. Fourthly, a subdivision
of only those first-stage samples is done for selecting the second-stage that are included in the
sample. Thus, we can conveniently divide the population into reasonably small sampling units.
Sometimes, a limitation is also associated with this method. Multistage sampling is less accurate
because of the involvement of several stages. Some researchers aver with confidence that a
single-stage process might be more accurate than a multistage process.
Methods of Unrestricted Random Sampling: In this method, each unit of the population has an
equal opportunity of getting selected in the sample. The personal bias of the market researcher
does not enter the selection process. That is because chance only determines which items
would be included in the sample.
Suppose that we have a universe with a total of 'N' elements. We want a simple random sample
of 'n' elements. Then, the following statements 'R' any of the following must be true:
(a) All the items of the sample are selected independently of one another.
(b) All the N items of the population have the same chance of being included in the sample.
(c) After each selection procedure, all the remaining units of the population have the same
chance of being selected for the sample.
If we make a sample in such a way that each unit selected from the population is returned
before selecting the next unit, then each item has a probability of 1/N for its selection. If the
selection is done in such a manner that each unit, which is selected, is not returned to the
population before making the next draw, then the probability of selection at first draw is 1/N, at
the second draw, 1/N-land at the third draw, l/(N-2) and so on. All possible samples of a given
size n are equally likely to be selected.
Unrestricted Random Sampling This method assumes that each site has an equal chance of
being part of the sample selected. Make a list of all project sites, perhaps by alphabetical order.
Every project site is given a number. Random sampling isn’t always the most convenient
method for choosing a sample.
In this method of unrestricted sampling, all the items of the universe are numbered, named on
separate slips of paper. These slips are folded (to ensure privacy and exclusion of sampling
bias). These are put in a drum. The drum is rotated and then, the required number of slips is
taken out of it at random. This method is popular in lottery draws, but paper slips should be of
same size shape and colors.
(ii) Tables of Random Numbers:
Random numbers are obtained by some mechanism which, when repeated many times, ensures
roughly equal frequencies for various combinations of numbers like 00, 01, 99, 999, etc.
Several tables of random numbers are available in markets. Tippet’s Random Number Tables
have 41,600 random digits. Fishers and Yates have developed a table of random numbers with
15,000 random digits. Rand Corporation has also issued a table of random members. These
numbers are selected by the researcher at random. Tippet’s numbers have been subjected to
many tests and investigations. Their randomness has been proved.
There are three advantages of the method of sample selection by random numbers. Firstly,
there is no personal bias of the researcher (or selector of numbers) because of the procedure of
random selection. Secondly, the random sample represents the universe in a better manner, if
we compare this method with judgment sampling. Thirdly, the researcher can check the
accuracy of his data because sampling errors follow the principle of chance.
This method has some limitations too. Firstly, we have to keep complete information about the
entire universe. Secondly, the size of the sample is very large as compared to that in stratified
sampling. Thirdly, the cost of collecting data is high because the population is normally living in
a large area. Fourthly, time for collecting data is very long.
Fifthly, random sampling by this method can produce many non-random results.
Question 4: Write short notes on
a) Design of a Research Project
The designing of a research project involves -
1. Stating a topic and problem, with appurtenant questions and/or hypotheses
2. Identifying the type of research called for in the topic
3. Learning more about the type of research selected
4. Organizing a library search for information related to the topic
5. Foreseeing needed data and their potential sources
6. Making a list of steps to be taken in conducting the study
7. Noting the procedures and tools required for collecting data
8. Anticipating the kinds of analyses the data will need
9. Drawing up conclusions from the data analysis
10. Selecting the appropriate format for reporting the research
The process includes -
The Research Question
It dictates the goal of the study, a summary of what is to be achieved, the methodology and the
type of analysis. Research Question should be Feasible, Ethical and Relevant.
Qualitative Research
Dictates the issues be clarified and the policies or practices that will be influenced.
Quantitative Research
Data Collection Techniques - Interviews, Surveys, Observation
Measurements and Outcomes
Analysis -
Data Cleaning, Descriptive Statistics, Advanced Analysis, Results
Reporting -
Tables and Figures, Interpretations, Scientific Writing
Proofreading
b) Ex post facto research
Ex post facto study or after-the-fact research is a category of research design in which the
investigation starts after the fact has occurred without interference from the researcher. The
majority of social research, in contexts in which it is not possible or acceptable to manipulate
the characteristics of human participants is based on ex post facto research designs.
This refers to those studies which investigate possible cause-and-effect relationships by
observing an existing condition or state of affairs and searching back in time for plausible
causal factors.
Despite studying facts that have already occurred, ex-post facto research shares with
experimental research design some of its basic logic of inquiry.
Example
A situation in which there has been a dramatic increase in the number of fatal road accidents
in a particular locality. An expert is called in to investigate.
Naturally, there is no way in which he can study the actual accidents.
However, he can study statistics, examine the accident spots, and ask the victims and
witnesses. So the expert will be in a position to identify possible determinants of the accidents.
These may include excessive speed, poor road conditions, careless driving, frustration,
inefficient vehicles, the effects of drugs or alcohol and so on.
On the basis of his examination, he can formulate hypotheses as to the likely causes and form
the recommendations. These may include improving road conditions, or lowering the speed
limit, or increasing police surveillance.
Therefore, in identifying the causes retrospectively, the expert adopts an ex-post facto
perspective.
c) Components of the Research Problem
1 - Research Consumer
There must be individuals or groups which have some difficulty or problem. The
individuals or the groups themselves may be researchers.
2 - Research-Consumer’s Objective
There must be some objectives to be attained and the research consumer must have some goal
he wants to achieve.
3 - Alternative Means to Meet the Objective
There must be alternative means or the courses of action for attaining an objective one wishes
to obtain.
4 - Doubt in Regard to the Selection of Alternatives
The existence of alternative courses of action is not enough. To experience a problem the
researcher must have some doubt as to which alternative to select. This means that research
must answer the question concerning the relative efficiency of the possible
alternative.
5 - There must be one or more Environments
There must be some environments to which the difficulty or problem pertains. A change in the
environment may produce or remove a problem. A researcher may have doubts as to which will
be the most efficient means in one environment but may entertain no such doubt in another.
Thus, a research problem is one which requires a researcher to find out the best solution for the
given problem so that the object can be attained optimally in the context of a given
environment.
d) Rephrasing the Research Problem
As a researcher one should rephrase the research problem. Once the following steps have been
followed,
1 - Understood the nature of the problem
2 - Studied the environment of the problem
3 - Discussed the problem with experts
4 - Surveyed and examined the available literature
Researcher needs to rephrase the problem into analytical and operational terms.
Points should be kept in mind while rephrasing a research problem
1 - Technical terms, words or phrases, with special meaning in a statement of the problem
should be clearly defined.
2 - Basic assumptions or postulates (if any) relating to the research problem should be clearly
stated.
3 - The scope of investigation on the limits within which the problem is to be studied must be
mentioned explicitly in defining the research problem.
Example
Initial research question
Why is productivity in Japan so much higher than in India?
Rephrased research question
What factors were responsible for the higher labor productivity of Japan’s manufacturing
industries during the decade 1971 to 1980 relative to India’s manufacturing industries.