Chapter: 3 - CENSUS AND SAMPLE METHODS OF COLLECTION OF DATA

There are 2,000 students in a college. An investigator wants to collect data regarding their family
background. He has two possible choices. First, he collects information relating to all the 2,000
students. Second, he collects information relating to some of the students (sample of students) who
would represent all the 2,000 students. In Statistics, the first approach for collecting data is called
Census Method and the second approach is called Sample Method. The present chapter focuses on
a comprehensive study of the census and sample methods of data collection.
1. CONCEPTS OF 'CENSUS' AND 'SAMPLE'
In Statistics, universe or population simply refers to an aggregate of items to be studied for an
investigation. To illustrate, there are 2,000 students in a particular college. If an investigation relates
to all the 2,000 students, then 2,000 would be taken as universe or population.
Each unit of these 2,000 is called Item. To further illustrate, 1 sugar mill out of the 10 sugar mills we
are studying, would be called an item. All the 10 sugar mills would constitute population or the
universe.
2. CENSUS METHOD:
METHOD OF COMPLETE ENUMERATION
Census method is that method in which data are collected covering every item of the universe or
population relating to the problem under investigation.
Suitability
Census method is suitable particularly for such statistical investigations which have
(i) small size of population,
(ii) widely diverse items in the population,
(iii) requirement of intensive examination of different items, and
(iv) High degree of accuracy and reliability.
Merits
Principal merits of census method are as under:
(1) Reliable and Accurate: Results based on census method are accurate and highly reliable. This is
because each and every item of the population is studied.
(2) Less Biased: Results based on census method are less biased.
It is because of the absence of investigator's discretion regarding the selection of sample items.
(3) Extensive Information: Information collected through the census method is quite exhaustive and
therefore, more meaningful because all the items of a universe are examined. For example,
population census in India gives exhaustive information relating to the number of people in different
parts of the country, their age and sex composition, education, status, occupation, and the like.
(4) Study of Diverse Characteristics: By using census method, one can study diverse characteristics
of the universe.
(5) Study of Complex Investigation: When items in a universe are of complex nature and it is
necessary to study each item, only census method can produce the desired results. Data on
country's population are collected by this method.
Census and Sample Methods of Collection
(6) Indirect Investigation: Census method can be successful, used in indirect investigations relating
to unemployment poverty, corruption, etc.
Demerits
However, there are certain demerits of census method as under:
(1) Costly: Census method is very costly and is, therefore, generally not used for ordinary
investigations. Only the Government or some big institutions can afford to use this method and that
too for specific purposes only.
(2) Large Manpower: Census method requires large manpower (enumerators). Training of a large
number of enumerators becomes essential, which is a very difficult process.
(3) Not Suitable for Large Investigations: If the universe comprises a large number of items, then it
may not be possible to cover each and every item. Census method becomes practically inoperative
in such situations.
3. SAMPLE METHOD
Sample method is that method in which data is collected about the sample on a group of items
taken from the population for examination and conclusions are drawn on their basis.
Sample method is widely used in our day-to-day life. A lady in the Kitchen, for example, tests only a
grain or two of the rice to know whether the rice is boiled or not. By examining only a few drops of
blood, a doctor determines the blood group of a person.
Suitability
Sample method is particularly suitable when:
(i) The size of population is very large,
(ii) Very high degree of accuracy is not needed,
(iii) Intensive examination of diverse items is not required, When different units of the universe are
broadly similar to each other.
Merits
Some of the principal merits of the sample method are as under:
(1) Economical: Sample method of investigation is economical because only some units of the
population are studied.
(2) Time Saving: In this method, only limited number of the items are investigated. As such the
process of investigation is time-saving, not time-consuming.
(3) Identification of Error: Because only a limited number of items are covered, errors can be easily
identified. To that extent sampling method shows better accuracy.
(4) Large Investigations: Sample method is more feasible in situations of large investigations than the
census method which generally involves unaffordable cost.
(5) Administrative Convenience: There is an administrative convenience in handling a limited
number of items. More capable and efficient investigators can be appointed.
(6) More Scientific: According to R. Fisher, Sample Method is more scientific because the sample
data can be conveniently investigated from various angles.
Demerits
Yet there are some demerits of the sample method as under:
(1) Partial: It is only a partial investigation of the universe. The investigator's bias in the selection of
the sample is not ruled out. Accordingly, the results may be biased as well.
(2) Wrong Conclusions: If the selected sample does not represent the characteristics of the universe,
the study may end up with wrong conclusions.
(3) Difficulty in Selecting Representative Sample: It is not very easy to select a sample which would
represent the characteristics of the entire population.
(4) Difficulty in Framing a Sample: Sometimes the universe may be so diverse that it becomes
difficult to frame a sample.
(5) Specialised Knowledge: Sampling involves a set of technical procedures. One must have the
technical knowledge of choosing a representative sample from the universe. Persons who are well-
versed with all the techniques of sampling are not easily available. Essentials of a Sample
In order to arrive at an unbiased and right conclusions, a sample must have the following qualities or
essentials:
(1) Representative: A sample must represent all the characteristics of the universe. It is possible only
when each unit of the universe stands equal chances of being selected in the sample.
(2) Independent: All units of a sample must be independent of each other. In other words, inclusion
of one item in the sample should not be dependent upon the inclusion of some other items of the
universe. Census
(3) Homogeneity: If more than one sample are selected from a universe, these samples should be
homogeneous (and no contradictory) to each other.
(4) Adequacy: The number of items in the sample should be fairly adequate so that some reliable
conclusions are drawn covering the characteristics of the universe as a whole.
METHODS OF SAMPLING
Following are the principal methods or techniques of sampling:
(1) Random Sampling
Random sampling is that method of sampling in which each and ever item of the universe has equal
chance of being selected in the sample. In other words, there is an equal probability for every item
of the universe being selected in the sample.
Which of the various items of the universe would get selected in the sample is beyond the control of
the investigator. The selection is left entirely to the chance factors. This method is used particularly
when various items of the universe are homogeneous or identical to each other. This method is
impartial and economical. Random Sampling may be done in any of the following ways:
(i) Lottery Method: In this method, paper-slips are made for each item of the universe. These slips
are shuffled in a box.
Then, impartially, some of the slips are drawn to form a sample of the universe.
(ii) Tables of Random Numbers: Some statisticians have prepared a set of tables called Tables of
Random Numbers.
A sample is framed with reference to these tables. Of all these tables Tippet's Table is most widely
used. Using 41,600 figures, Tippet has involved 10,400 numbers comprising of four units each. For
the use of this method, all items of the universe are first arranged in an order. Then using Tippet's
Table the required number of items are selected as are needed for a sample.
Merits
Following points may be noted on the merits of random sampling:
(i) This method is free from personal bias of the investigator.
(i) Each and every item of the universe stands equal chances of being selected.
(iii) The universe gets fairly represented by the sample.
(iv) This is a very simple and straightforward method.

Demerits
However, there are two notable demerits of this method. These are as under:
(i) This method does not guarantee proportionate representation of different items in the universe.
ii) Random sampling does not give weightage to certain important items in the universe.
(2) Non-Random Sampling
Non-random sampling includes all those methods of sampling in which all the units of population do
not have equal probability (or equal chances) of being selected in the sample of study. Instead,
convenience of the investigator or his personal judgement plays an important role in the selection of
a sample. Some of the well-known methods of non-random sampling are as under:
(i) Purposive or Deliberate Sampling: Purposive sampling is that method in which the investigator
himself makes the choice of the sample items which in his opinion are the best representative of the
universe.
Thus, in this method of sampling, selection of the sample items is not left to the chance factors; it is
simply made by choice. This method of sampling is specifically suitable when some of the items in
the universe are of special significance and ought to be included in the sample. For example, if an
investigation is to be made relating to the iron and steel industry in India, then the inclusion of such
industries as the Tata Iron and Steel Company is obviously a purposive selection. However, there is a
considerable possibility of personal bias in purposive sampling. As a result, it loses its credibility.
Merits
(a) This method is flexible to allow the inclusion of those items in the sample which are of special
significance.
(b) Selection of items can be deliberately tuned to the purpose of study.
(c) It is a very simple technique of selection of the sample items.
Demerits
(a) There is a possibility of personal bias in the selection of items.
(b) Because of the possibilities of personal bias, reliability of the results becomes doubtful.
(ii) Stratified or Mixed Sampling: This method of sampling is generally adopted when population
consists of different groups with different characteristics. According to this method of sampling,
population is divided into different strata having different characteristics and some of the items are
selected from each strata, so that the entire population gets represented. Each stratum should be
represented in correct proportion in the sample.
To illustrate, suppose there are 50 students in Class XI. Out of them, 30 have studied Mathematics
and 20 have studied
Home Science in their Class X. Thus, the population of the 50 students gets divided into two strata
consisting of 30 and
20 students respectively. From each of these strata, items would be selected proportionately such
that the sample represents the characteristics of the entire population. If, of the total 50 students,
only a sample of five is to be selected, then we shall randomly select three out of the first stratum
(comprising 30 students) and two out of the second stratum (comprising 20 students). However,
items may also be selected proportionately from different strata.
An investigator may prefer to select four students from amongst those who studied Mathematics
and only one from amongst those who studied Home Science. The choice will be governed by the
nature of the enquiry and wisdom of the investigator.
Stratified Sampling is also called mixed technique of sampling because this method involves the
mixture of both purposive sampling and random sampling. The division of population into different
strata is purposely done while selection of the items is done at random.
Merits
(a) This method covers diverse characteristics of the population.
(b) On the basis of diverse characteristics of the population, a comparative analysis of the data
becomes possible.
(c) This method of sampling offers reliable as well as meaningful results.
The Principal Merit of Stratified Sampling
It allows selection of such items which represent diverse characteristics of the population.
Demerits
(a) This method is suitable only when there is a complete knowledge about the diverse
characteristics of the population. Therefore, this has a limited scope.
(b) There is a possibility of bias at the time of classification of the population into different strata.
(c) When the size of population is already small, it may be difficult to further divide it into smaller
parts/strata. iii) Systematic Sampling: According to this method, units of the population are
numerically, geographically and alphabetically arranged. Every nth item of the numbered items is
selected as a sample item. To illustrate, if 10 out of 100 students are to be selected for a sample,
then 100 students would be numbered and systematically arranged. One item of the first 10 would
be selected at random. Subsequently, every 10th item from
the selected number will be selected to frame a sample. If the first selected number is 5th item, then
the subsequent! Numbers would be 15th, 25th, 35th, 45th, 55th, 65th, 75th 85th and 95th. This
method of sampling is, in fact, a short Cut method of Random Sampling.
Merits
(a) This is a very simple method. Sample is easily determined.
(b) There is hardly any possibility of personal bias in this method.
Demerits
(a) Every item in the population does not get equal chance of being selected because only the first
item is selected on the basis of random sampling.
(b) If all the items in the population are homogeneous, this method of sampling serves no specific
purpose.
(iv) Quota Sampling: In this method, the population is divided into different groups or classes
according to different characteristics of the population. Some percentage of different groups in total
population is fixed. Further, some quota of the items to be selected as sample-items is fixed for each
group. The investigator selects the fixed number of items from each group to frame a sample.
This method of sampling is not very expensive. But there is a high possibility of personal bias at the
time of selection of the items by the investigator. Accordingly, the reliability of results becomes
questionable.
(v) Convenience Sampling: In this method, sampling is done by the investigator in such a manner
that suits his convenience. To illustrate, an investigator may select a sample of teachers merely by
referring to the college prospectus. This method is the simplest and least expensive, but unscientific
and unreliable. It depends too much on the whims of enumerators.
Reliability of Sampling Data
The reliability of the sampling data means that the characteristics of the universe are fully
represented by the sample. It depends mainly on the following factors:
(1) Size of the Sample: Reliability of sampling
depends on
the size of the sample. If its size is very small, it will fail to represent the population. Accordingly, the
conclusions would lack reliability.
(2) Method of Sampling: If the method of sampling is not simple and exhaustive, it will not
adequately represent the population. Consequently, the results will not be dependable.
(3) Bias of Correspondents and Enumerators: Personal bias of the correspondents and enumerators
should be as less as possible. Otherwise, reliability of the sampling data is bound to suffer.
(4) Training of Enumerators: Reliability of sample also depends upon the training of the
investigators. If they are not trained to make them expert in their field of investigation, the sample
will lack reliability.
Census and Sampling Methods: A Comparative Look
Some of the principal differences between census and sampling methods are as under:
(1) Coverage: In the census technique, an investigator collects information relating to all the items in
the population. In sampling method, on the other hand, only some of the items which represent the
population are covered for an investigation.
(2) Suitability: Census method is suitable when the area of investigation is relatively small. On the
other hand, when the area of investigation is large, it is the sampling method which is generally
used.
(3)
Accuracy: There is generally a greater degree of accuracy in the results based on the census method
than the sampling method. This is because in the census method each and every item of the
population is studied. As against it, there is less accuracy and reliability in the sampling method
because it studies only a few items of population. However, errors can be easily detected and
removed in the sampling method because of the small number of items. To that extent sampling
method shows greater accuracy than the census method.
(4) Cost: Sampling method is certainly much less expensive than the census method. Smaller the
sample size out of the given population, lesser the cost of investigation.
(5) Time: Sample method is less time consuming than the census method.
(6) Nature of Items: Census method is particularly suitable when the items in the population have
diverse characteristics. On
Census and Sample Methods of Collection of
the other hand, sample method is suitable when items in the population are homogeneous.
(7) Verification: Verification of the statistical information obtained through census method is
generally not possible.
It would involve huge expenses and the repetition of the whole process. Sample information, on the
other hand, can be easily verified. In case of doubt, enumeration can be done again and facts
verified accordingly.
In short, sample method of statistical investigation is generally preferred to the census method
because the former is less expensive in terms of the time, money and efforts involved.
However, for the successful application of the sample method, it is very essential that the sample
items represent the characteristics of population as a whole.