Status of Statistics

For the last few centuries, statistics has remained a part of mathematics as the original
work was done by mathematicians like Pascal (1623-1662), James Bernoulli (1654-1705),
De Moivre (1667-1754), Laplace (1749-1827), Gauss (1777-1855), Lagrange, Bayes,
Markoff, Euler etc. These mathematicians were mainly interested in the development of
the theory of probability as applied to the theory of games and other chance phenomena.
Till early nineteenth century, statistics was mainly concerned with official statistics needed
for the collection of information on revenue, population and area of land under cultivation
etc. of a state or kingdom.
The science of statistics developed gradually and its field of application widened day
by day. Hence, it is difficult to give an exact definition of statistics. The definition changed
from time to time depending upon its use and application. Numerous definitions have
been coined by different people. These definitions reflect the statistical angle and field of
activity. A few of these definitions are given below:
The word ‘statistics’ is known to have been used for the first time in “Elements of
Universal Erudition” by Baron J.F. Von Bielfeld, translated by W. Hooper M.D. (3 vols.,
London, 1970). Here statistics is defined as:
“The science that teaches us what is the political arrangement of all modern states
of the known world.”
Webster “Theclassified facts representing the condition of the people in a state, especially
those facts which can be stated in numbers or in tables of numbers or in any tabular or
classified arrangement.”
Horace Secrist “Statistics is an aggregate of facts affected to a marked extent by the
multiplicity of causes, numerically expressed, enumerated or estimated according to a
reasonable standard of accuracy, collected in a systematic manner for a predetermined
purpose and placed in relation to each other.”
Professor A.L. Bowley gave several definitions of statistics as:
(i) The science of counting.
(ii) The science of averages.
(iii) The science of measurement of social phenomena, regarded as a whole in all its
manifestations.
Status of StatisticsStatus of Statistics
1Chapter

2 Basic Statistics
(iv) A subject not confined to any one science.
A.L. Boddington “Statistics is the science of estimates and probabilities.”
Croxton and Cowden “Statistics may be defined as the collection, presentation, analysis
and interpretation of numerical data.”
Wallis and Roberts “Statistics may rightly be regarded as a body of methods for making
wise decisions in the face of uncertainty.”
R.A. Fisher “The science of statistics is essentially a branch of applied mathematics
and may be regarded as mathematics applied to observational data.”
Of all the definitions, the one given by Fisher is considered to be most exact. Fisher’s
definition is most exact in the sense that this covers all aspects and fields of statistics.
In view of the latest developments in the field of statistics, it is considered as a science
of decision making, under uncertainty, with or without data.
The credit of widening the use and scope of statistics mainly goes to the statisticians
of England. The concept and field of statistics in the twentieth century have changed
totally.
Further, the theory of inference, design of experiments and sampling theory have
proved landmarks in the development of statistics. Some other contributions further support
this point of view. Francis Galton invented the regression theory and pioneered the use of
statistical methods in biometry. Karl Pearson developed the theory of distribution and
correlation analysis. His invention of chi-square test brought statistics to lime light. W.S.
Gosset developed t-test in 1908 and called it the student's t-test. R.A. Fisher did a lot of
work in various directions namely, the theory of estimation, the fiducial inference, exact
sampling distributions, the theory of design of experiments and testing of hypothesis.
Fisher used statistical methods in a number of sciences like biometry, agriculture, genetics,
sociology and education. His concepts were fundamental and oriented to the application
side. Since then the field of statistics has been widening day by day and is being used
increasingly in sociometry, psychometry, biometrics, and technometrics etc.
The statistical methods or techniques are applicable only when some data are available
irrespective of the method of data collection. The data can be quantitative as well as
qualitative. If the data are qualitative, they are quantified by using techniques like ranking,
scoring, scaling or coding etc. The data are collected either by experiments or by survey
methods (directly or indirectly) and they are tabulated and analysed statistically. Whatever
may be the resulting value(s) obtained from analysis, proper and correct inferences have
to be drawn from these numerical values. These inferences lead to a final decision.
On the basis of these ideas, we can broadly give the following functions of statistics:
(i) Collection of data
(ii) Tabulation of data
(iii) Analysis of data
(iv) Interpretation of results.
The four functions given above will be described adequately later. Further, their
application and utility will obviously be clear from the discussion of the subject matter
given in the body of this book.

Status of Statistics 3
COLLECTION OF DATA
Once it is decided what type of study is to be made, it becomes necessary to collect information
about the concerned study, mostly in the form of data. For this, information has to be
collected from certain individuals directly or indirectly. Such a technique is known as
survey method. These are commonly used in social sciences i.e., the problems relating to
sociology, political science, psychology and various economic studies. In surveys, the required
information is supplied by the individual under study or is based on measurements of
certain units. Generally, the respondents or units are selected from a population using
some standard sampling techniques. Another way of collecting data is by experimentation
i.e., an actual experiment is conducted on certain individuals or units about which the
inference is to be drawn. Such experimental studies are common in agriculture, biology,
medical science, chemistry, industry etc.
TYPES OF DATA
There are two categories of data namely, (i) primary data and (ii) secondary data.
Primary Data
The data, which are collected from the units or individual respondents directly for the
purpose of certain study or information, are known as primary data. For instance, an
enquiry is made from each tax payer in a city to obtain their opinion about the tax collecting
machinery. The data obtained in a study by the investigator are termed as primary data.
If an experiment is conducted to know the effect of certain fertilizer doses on the yield or
the effect of a drug on the patients, the observations taken on each plot or patient constitute
the primary data. Hundreds of such examples can be cited.
Secondary Data
The data, which had been collected by certain people or agency, and statistically
treated. Now the information contained in it is used again from records, processed and
statistically analysed to extract some information for other purpose, is termed as secondary
data. For instance, if the data given in different census years is again processed to obtain
trends of population growth, professional changes, changes in sex ratio, mortality rate
etc., it is termed as secondary data. Usually, secondary data is obtained from year books,
census reports, survey reports, official records or reported experimental findings. Different
organizations and government agencies publish information (data) in the form of reports,
periodicals, journals etc. Names of organizations and their publications with relevant details
are given in the chapter on Statistical Organization in India (Chapter 22).
PROCESSING OF DATA
Before tabulation of primary data, it should be scrutinized for (i) completeness,
(ii) consistency, (iii) accuracy and (iv) editing.
Completeness
If the answer to some important question in a schedule or questionnaire is missing, it
becomes necessary to contact the informant again and complete the missing information
in the report. In case, such an information cannot be completed, the schedule or
questionnaire should be discarded or revised.

4 Basic Statistics
Consistency
Some information given by the respondent may not be compatible in the sense that
an information furnished by the individual either does not justify some other information
or is contradictory to earlier one. For example, the total expenditure exceeds the income
reported by the respondent, the number of children mentioned is less than total number of
sons and daughters, then the respondent should again be contacted to rectify the mistake.
Accuracy
It is of vital importance. If the data are inaccurate, the conclusions drawn from it
have no relevance or reliability. By checking the schedules or questionnaire only a little
improvement can be made e.g., if the sum of certain figures is wrong, it can be corrected.
But if the investigator has either made a false report or the respondent has deliberately
supplied wrong information about his income, age or assets etc., editing will be of no use.
In recent times, checks have been evolved to attain accuracy e.g., by sending supervisors
to check the work of investigators or reinvestigating a few respondents after a certain gap
of time.
Editing
To maintain homogeneity, the information sheets are checked to see whether the
unit of information or measurement is the same in all the schedules. For instance, some
people might have reported income per month and some annual income. In such a situation,
it has to be converted to the same unit during editing. It should also be checked whether or
not the same information has been supplied for a particular question in all the information
sheets. The ambiguity arises due to various interpretations of the same question and should
be removed.
Once the primary data have undergone the above four processes it is fit for further
analysis.
Large scale data cannot be collected repeatedly because of the paucity of time, money
and personnel. Hence the use of secondary data for certain studies is inevitable. While
making use of the secondary data, one should always take care of the following points.
(a) One should see whether the data are suitable for study.
(b) The source of data should also be viewed, keeping in mind whether at any time,
it is reliable or not. If there is any doubt about the reliability of data, it should not
be used.
(c) It should be noted that the data is not obsolete.
(d) In case the data are based on a sample, one should see whether the sample is a
proper representative of the population.
(e) The primary data has been handled carefully by skilled persons only.
Once the above points are observed in the secondary data, it is ready to be used for
further analysis.
ACCURACY OF MEASUREMENT
Accuracy of a measurement depends on the precision required, the tools available for the
measurement and the skill of the person undertaking the task. Though absolute accuracy
is neither possible nor desired in statistics, a reasonable degree of accuracy is a must. For

instance, the monthly income of a person should be rounded to the nearest of ten of rupees,
height is measured upto one-tenth of a cm and calculation of age in years and months is
fairly accurate.
ROUNDING OF FIGURES
Sometimes the figures (values) are rounded by reducing one or more decimal places to the
unit place or nearest to the ten or hundredth of a number. The universal rules of rounding
are – if the decimal place value to be rounded is less than 5, it should be deleted straightway,
and if greater than 5, the preceding number is increased by 1. In case the decimal place
value to be rounded off is 5, the rule is to delete it if the preceding number is even and
increase the preceding number by one if it is odd.
For instance, 9.74 will be rounded to one decimal place as 9.7 and 9.78 will be rounded
as 9.8. Moreover, 9.75 will be rounded as 9.8 and 9.85 will also be rounded as 9.8 according
to the rule. The rule is applicable in general for any numerical value.
ABSOLUTE AND RELATIVE ERROR
Before giving absolute and relative error it is worth pointing out that in statistics an error
is different from a mistake. The mistakes in counting, weighing, measuring or reporting
are not errors in statistical sense and should be regarded as mistakes only. Error in statistical
sense means the difference between the actual value and the value under consideration,
generally an estimated value. An absolute error (A.E.) is the absolute difference between
the actual value (X) and its estimated value (x).
A.E. = |X – x| ...(1.1)
The relative error (R.E.) is the ratio of the absolute error to the actual value. i.e.,
R.E. =
X
xX || −
...(1.2)
Very often the relative error is given in per cent, i.e.,
R.E. = 100
||
×
−
X
xX
...(1.2.1)
The smaller the relative error, the better is the result.
METHODS OF ENQUIRY
The objective of a study, the population under study, the units or individuals from whom
the information is to be collected is ascertained first. Then keeping in view the purpose
and importance of the investigation and types of respondents, the statistician has to choose
one of the three methods of enquiry given below:
1. Personal enquiry method.
2. Correspondence, i.e., mailed questionnaire method.
3. Direct observational method.
Personal Enquiry Method
Before starting the investigation, a question sheet is prepared which is called schedule.
The schedule contains all the questions which would extract a complete information from
a respondent. Often, the schedule contains the likely answers also. As a precaution in any

6 Basic Statistics
method a few schedules are tested by filling them before the commencement of the survey.
This pre-testing of schedules removes certain discrepancies like ambiguity of the questions
and irrelevant questions in the schedule. Such pre-testing is known as a pilot survey.
In this type of enquiry, the investigator contacts the respondent personally and asks
him questions given in the schedule one by one and notes down his replies on the schedule.
If the unit is a household, the investigator contacts the head of the family to fill up the
schedule. Such a practice is called a direct personal interview method.
Sometimes, the information is not collected directly from the respondent but from a
third person who is expected to know him well. Such an approach is useful in case where
the respondent is expected to conceal information about himself. For instance, a person
who is addicted to alcoholic drinks or suffering from some infectious disease may hide
correct information about certain things. Hence, an indirect enquiry will give better
information. Such a method of enquiry is known as indirect personal enquiry method.
The main advantage of this method is that the information is likely to be complete, more
correct and some additional information can be retrieved. The person may be persuaded to
reply to some of the more personal questions too.
Mailed Questionnaire Method
When a survey is spread over a vast area and the respondents are educated, a mailed
questionnaire method is preferred in comparison to the personal interview method. In this
method, a questionnaire is mailed to each and every respondent. They are requested to fill
up the questionnaire and send it back. This method is less costly and less time consuming.
The difficulty in this method is that a large number of respondents do not return the duly
filled in questionnaires.
Direct Observational Method
In this approach an investigator stays at the place of survey. He does not make
enquiries but notes down the observations himself. For instance, to understand the
migratory habits of a variety of birds, the investigator stays at the birds sanctuary and
notes down the movements to and fro every day. This method of survey is not used
frequently.
TABULATION OF DATA
The information collected through an enquiry or experiment may be presented in the form
of tables or graphs. These tables enable us to come to some conclusion and make certain
comparisons. Besides this, they set a stage fit for analysing the data. A table is a systematic
arrangement of data in rows and columns, which is easy to understand and makes data fit
for further analysis and drawing conclusions. On the other hand, a diagram, a chart or a
graph is the visual form for presentation of data. Such a presentation makes it easy for a
common man to understand the fluctuations, variations and the existing state of affairs of
a phenomenon. The diagrams, charts or graphs may be two or three dimensional. The
charts containing pictures with captions are called pictograms. These are discussed in
Chapter 2.
Regarding the discussion on tabulation of data, a table in general consists of the
following parts.

Title
It gives information about the contents of the table. In a two-way table captioning
levels the data to be presented in the columns whereas stub levels the data to be presented
in the row.
Body of the Table
It contains the numerical information pertaining to the various factors or characters.
Footnote
It is a statement which gives some specific information about the contents given in
the body of the table.
Source Note
It gives information about the source of data given in the table if it has not been
collected by the person presenting them.
Keeping in view the above mentioned features of a table, a number of tables having
varied classification can be constructed. Some tables with their methods of construction
are discussed in Chapter 2.
ANALYSIS OF DATA
As given by Zacks1 “one of the most important objectives in the primary stage of statistical
analysis is to process the observed data and transform it to a form most suitable for decision
making. This primary data processing generally reduces the size of the original sets of
sample values to a relatively small number of statistics. It is desired, however, that no
information relevant to the decision process will be lost in this primary data reduction.”
The measures of central tendency and dispersion like mean, median, mode, mean
deviation and standard deviation etc., are parts of analysis of data along with estimation
and testing of hypothesis. Besides these, any other statistical tools used or operations done
for drawing inference or making decision on the basis of data, are known as analysis of
data.
Note: For any statistical analysis on computer, one needs to create a data file. Procedure of creating
data in SPSS V 11.5 has been given at the end of this chapter.
INTERPRETATION OF DATA
Once the data have been analysed, some numerical value(s) which gives information partly
or wholly about the population under study can be achieved. The main job consists of
attaching physical meaning and giving interpretation to the numerical results useful in
real life. This must be true in its meaning and sense. The quality of interpretation depends
more and more on the experience and insight of the person. No pre-conceived ideas should
be thrusted on the numerical results obtained out of analysis of data. Also no attempts
should be made to draw more inferences than the results are actually liable to.
Especially spuriously calculated values should have no conclusions. For example, if a
yearly data for the last decade regarding the production of steel and production of shoes is
1. Zacks, S., The Theory of Statistical Inference, John Wiley & Sons, New York, 1971, p-29.

8 Basic Statistics
taken and one finds a significant positive correlation2 between the production of steel and
production of shoes, even then it should never be interpreted that production of shoes
increases in proportion to the production of steel. These two items are in no way related
with regard to their production. Hence, such a correlation is spurious and carries no sense.
LIMITATIONS OF STATISTICS
1. The main limitation of statistics is that it cannot deal with a single observation or
value.
2. Statistical methods are not applicable to studies which measure qualitative
characters and cannot be coded in numerical value.
3. The statistical study does not take care of the changes occurring to the individuals.
But it does reveal the changes occurring in a mass or group of individuals. For
instance, per capita gross national product (GNP) at factor cost in 1970 is
Rs. 673.80 and in 1980 it is Rs. 1690.30. On the basis of these figures it can
definitely be concluded that per capital GNP has increased by more than Rs.
1000 but it must be kept in mind that it gives no information about an individual.
4. Statistical statements or conclusions are generally not true or applicable to
individuals but are applicable to the majority of cases. For instance, the statistical
information says that 90 per cent patients of a particular disease die during
operation. If a doctor has performed nine operations so far and all the nine patients
have died during operation, he still cannot assure the tenth patient that he will
be cured after the operation.
There can be hundreds of situations which can be enunciated under the heading of
limitations of statistics. These ideas are of common sense and can be thought of by
individuals themselves.
DISTRUST AND MISUSE OF STATISTICS
Some people think that the inferences based on statistical data are very reliable as the
numbers cannot be false while others do not trust statistical results at all. According to the
non-believers these numerals are the tissues of falsehood as Disreali said ‘‘There are three
degrees of lies—lies, damned lies and statistics.’’ Darrel Huff 3 said ‘‘A well wrapped statistics
is better than Hitler’s ‘big lie’, it misleads, yet it cannot be pinned on you.’’ Another statement
by Huff4 showing his distrust in statistics is “ The secret language of statistics, so appealing
in a fact minded culture, is employed to sensationalise, inflate, confuse and oversimplify.”
All this confusion has emerged due to the fact that it is not easy to differentiate between
fictitious data and the data collected and tabulated in a well planned manner.
Statistics is misused either deliberately or often due to the lack of knowledge. As
defined earlier statistics is a tool, technique or an approach to deal with the data to arrive
at correct conclusion or decision. If intelligibly used, it gives wonderfully good results and
if misused, it can be disastrous. This can be very well supported by the statement, “Figures
2. Correlation coefficient has been discussed in Chapter 15.
3. Darrell Huff, How to lie with statistics. p. 9.
4. Ibid., p. 8.

won’t lie, but liars figure.” Another statement in support of the misuse of statistics is
‘‘Statistics is like clay, of which you can make a god or a devil.’’
The conclusion one comes to is that at present times, statistics is a highly developed
science with a deep rooted mathematical base. It is applicable to a large number of social,
economic and business phenomena. Statistics is the backbone of industrial research, basic
science research and planning.
QUESTIONS AND EXERCISES
1. Give three definitions of statistics which you consider most appropriate.
2. Statistics is not a science, it is a scientific method. Examine this statement and give
your views on it.
3. Explain clearly the functions and limitations of statistics.
4. Give various methods of collection of data. Describe each of them adequately.
5. Differentiate between the following:
(a) Primary data and secondary data.
(b) Absolute error and relative error.
6. Comment on the following statements:
(a) Statistics is an aid, but not a substitute for commonsense.
(b) Figures do not lie.
(c) With statistics anything can be proved.
7. Write a note on the misuse, limitations and distrust of statistics.
8. What operations should be performed before the data is used for analysis?
9. What are the drawbacks with which statistics suffers? Give two examples of the misuse
of statistics. (B.Com., Raj. 1962)
10. Explain various methods in the collection of statistical data. Of these, which would you
choose? Give reasons. (M.A., Vikram, 1963)
11. What is a statistics? How far do you think that the knowledge of statistics is essential
in the study of economics? (M.A., Lucknow, 1963)
12. Write, in brief, a history of the growth and development of the science of statistics
pointing out specially the landmarks. (M.A., Kolkata, 1952)
SUGGESTED READING
Harvey, J.M. (1969). Sources of Statistics, Clive Bingley.
McCarthy, P.J. (1957). Introduction to Statistical Reasoning, McGraw-Hill Book
Company, New York.
Monroney, M.J. (1956). Facts from Figures, Penguin Books, Baltimore.
Reichman, W.J. (1961). Use and Abuse of Statistics, Penguin Books, Baltimore.
Simpson, G. and Kafka, F. (1971). Basic Statistics, 3rd ed. Oxford and IBH, Kolkata.
Snderson, T. and Sclove, S. (1978). An Introduction to the Statistical Analysis of Data,
Houghton Mifflin, Boston.

10 Basic Statistics
APPENDIX
PROCEDURE FOR CREATING DATA FILE IN SPSS V.11.5.
When you start SPSS, it looks like any other spreadsheet as given below. For example, in
the following window also known as data editor, you can have a feel of Ms-Excel. Any
version of SPSS data editor has a grid of rows and column for data entry.
Window 1: SPSS Data Editor
Bottom of the window shown above has two panes i.e., Data View and Variable View.
Before the entry of data, users must create the variables in the Variable View. For example,
before creating data sheet for year wise export figures, both the variables need to be
defined as shown below.
Window 2: SPSS Data Editor–Variable View

As it can be seen, first column is the name of variable. Second column denotes the
data type. Default type is Numeric but when you click on this cell, following window
appears where exact data type can be selected. While other types are self-explanatory, for
character data, user should select String data type. Specification about each data type can
be obtained from the Help button in the window. It is noteworthy to mention about the
last column Measure where you can choose Nominal, Ordinal or any other Scale through
the checkbox.
Window 3: Variable Type
After defining the variables, click on Data View in the bottom and enter data as
given below. (Refer Example 2.7)

12 Basic Statistics
After entering the data, click on File menu and Save option for saving the file.
Procedure for saving the file is the same as in any other software. Saved file automatically
gets an extension of .Sav. You can also use already existing database created in normal
spreadsheet, DBMS, text processor or even the statistical software Systat or SAS. For this,
click on File menu and choose Open. In this Window, data type needs to be changed in
the bottom.
Using SPSS, wide gamut of statistical analysis can be performed as given at the end
of each chapter. Once you get the output for any analysis, you can save the file in the
normal way. SPSS output file automatically gets an extension of .spo.

Status of Statistics

Recommended

More Related Content

What's hot (20)

Similar to Status of Statistics (20)

Recently uploaded (20)

Status of Statistics