Statistical education

1
D. S. HOODA
Jaypee Institute of Engineering & Technology
A.B. Road, Raghogarh-473226
Distt. Guna- M.P.
INDIA
ABSTRACT
In the present talk a brief account of Indian Statistical
System and contributions of Indians to development of
statistics is presented. Statistics education at schools and
undergraduate levels is discussed in detail. Educational and
research activities in statistics at premier institutions and
universities are highlighted. In the end some suggestions for
future planning at national and international levels in
statistics education are made for discussion.
STATISTICS EDUCATION IN INDIA: A REVIEW

2
INTRODUCTION
Statistics as a subject is just a century old science. During
last few decades it has penetrated into almost all sciences like
agriculture, biology, business, social, engineering, medical sciences,
etc. Its wide and varied applications have lead to the growth of many
branches, such as Industrial Statistics, Biometrics, Business
Statistics, Bio-statistics and Agricultural Statistics. These branches
have emerged as distinct entities or subjects with a bulk of statistical
techniques specific to their application areas. Agricultural Statistics
is that area of statistical science which deals directly with the
problems of field experimentation and interpretation of results in
agricultural sciences. It was R.A. Fisher whose dedication and
continued efforts put the statistical science on a strong mathematical
foundation. The theory of experimental designs and other statistical
techniques developed by Fisher constitute the backbone of the
Agricultural Statistics.

3
In advanced countries with established
scientific tradition there is continuing concern
with validity of data and validity of conclusions,
while in underdeveloped countries the principle
of authority is still dominant and the question of
validity scarcely arise.
Statistics, therefore, remains a matter of formal
or administrative sanctions as anything having
the official stamp must be accepted as
authoritative.

The question of comparability or compellability
does not arise. But the scenario has changed after
globalization and industrialization.
Now the need of statistics of increasing coverage
and accuracy becomes more and more insistent.
India has already reached this stage, where the
statistical system in other developing countries is
unable to meet the need.
4

5
Prof. P.C. Mahalanosis in 1965 stated :
“ Indian statisticians have high reputation in the
world and yet the statistical system in India is
weak and lacking behind in the timely flow of
information required for social and economic
affaires. Why?”
The future of statistics needs to be discussed in
view of recent development in information
technology such as data mining, data
communication and information processing
networks and requirement of end users.

6
INDIAN STATISTICAL SYSTEM
Collection and use of statistics for administrative
purposes in India has a long history spread over many centuries.
The Arthasastra and the Ain Akbari mention the practice of
numerical data collection for purposes of statecraft in ancient and
medieval India.
The Mughals had a system of collection and compilation of
crop statistics to help them in land revenue collection. Later on,
Britishers created their own data generating system to serve their
specific ends.
During the British period, consolidated efforts were made for
the collection of socio-economic data. Their system was restricted
to a few specific fields like trade and commerce, selected
industrial products, population, some basic crop statistics and
livestock.

7
The Indian Statistical System (ISS) is one of the
largest institutional framework having a vast
wealth of information, not all of which gets due
attention of decision-makers. It is one of the oldest
and the largest network supported with adequate
facilities for data management by competent
personnel. The system claims a wide coverage of
information items, area and population. When we
talk about a national statistical system, we can
reasonably ask whether such a system possesses
the four characteristics of a system viz. content,
structure, communication and control?

Such a system should be able to:
(i) Identify its long-term and short-term
objectives as well as a strategy to achieve
these objectives.
(ii) Ensure a structure in terms of job-
differentiation by recognizing different types
of information to be collected and presented the
different aspects of information processing for
which the responsibility and authority should
be clearly vested, possibly in different groups
of people.
8

9
(iii) Establish clear lines of communication
within itself to take care of inter-group or
collective responsibilities as well as with the
environment.
(iv) Review its contents, structure and
communication regarding their effectiveness as
well as efficiency and modify or control one or
more of these as and when found necessary.

There is gap between theory and practice. There is
gap between the means and the end in the absence of
any clearly perceived purpose.
There is lack of appreciation of the need of cross-
examining the data, although it is the first
responsibility of a statistician.
Unfortunately, collection and scrutiny of primary
information or processing and handing of data were
usually not liked by government officials and statistics
was, therefore, not considered a popular subject for
acquiring skill and experience in pre-independence
days. 10

11
Just after independence in 1947, the system
of data collection followed by the Britishers was
found inadequate to meet the necessity of a strong
database covering a variety of social and economical
aspects.
The existing system even did not provide the basic
data required for estimation of national income,
which is essential for assessing performance and
progress of the economy. The immediate task,
therefore, was to set up a statistical system capable of
filling the large gaps in the data essential for
formulating economic plans.

A very important step in this direction was the
creation of the Directorate of National Sample
Survey in 1950.
Its aim was to collect essential statistics related
to the socio-economic conditions and agricultural
production in India. Due to the sustained efforts of
academicians and official statisticians the Indian
Statistical System have attained the height of the
most comprehensive statistical system of a
developing country.
12

13
To-day India has a decentralized statistical
system between centre and the states. The
demarcation has been done partly on the functional
basis and partly on the regional basis. Statistics of
items like foreign trade, banking and currency and
census are wholly allocated to the centre and that of
like agriculture and education are assigned to the
States. There is also a common category of items
known as concurrent list which both the centre and
the States can operate simultaneously to meet their
respective requirements. The department of
Statistics (DOS) of the Central Government is the
apex body in the official statistics of India.

The ISS pertains to the collection,
compilation and dissemination of data relating
to socio-economic, agricultural and industrial
statistics in India.
Ministry of Agriculture, the Central
Statistical Organization (CSO) and the
National Sample Survey organization (NSSO)
are important agencies at the national level
involved in collection, compilation and
dissemination of data.
14

15
The CSO is mainly responsible for coordination of
statistical activities as well as maintaining statistical
standards. The NSSO has been a leading sample survey
organization since its establishment in 1950 and continues to
conduct major multi-subject surveys that provide valuable
data required by the policy makers.
The NSSO conducts large-scale surveys at the national
level and collects and disseminates information on different
areas. The NSSO, under the scheme of improvement of crop
statistics, also provides technical guidance to the states in
respect of the crop estimation surveys and performs sample
checks to assess the quality of primary work done by the
state agencies in area and crops estimation surveys.

Department of Planning and Statistics,
Ministry of Agricultural and some other
Government Departments, institutions and
autonomous bodies and non-government
organizations are actively involved in the
creation of large database.
Many of them compile and publish information
on agricultural Statistics and livestock
population collected through periodic censuses.
16

17
CONTRIBUTION OF INDIANS TO STATISTICS
Indian teachers, researchers and practitioners have
played a significant role in development of statistics as a
subject. Many of them have shifted to the advanced
countries like USA and enriched the science of statistics
and have indeed left an indelible mark in the statistical
literature. Consequently, the current status of statistics in
India could not develop and ‘ Brain-Drain’ was the main
reason for its deterioration as most of the eminent
statisticians have shifted to the advanced countries. This
had caused a detrimental effect on statistical education and
research in India. However, now the situation has changed
drastically.

18
Indian statistician have made tremendous contributions:
D Basu Bayesian Statistics
V.S. Huzurbazar : Bayesian inference properties of the exponential
family of distributions and the uniqueness of the likelihood equation.
D.B. Lahiri: Lahiri’s Method of sample selection with PPS
P.C. Mahalanobis: Concept of optimum survey design, Pilot survey, D2
Interpenetrating network of samples
C.R. Rao: Cramer-Rao Lower bound, Rao- Blackwell theorem , First
and second order efficiency , g-inverse, growth curves, Unified theory of
linear models, etc..
R. Masani: Wierner-Masani theory of Multivariate Stationary Stochastic
Processes.
P.V. Sukhatme: A pioneer in Agricultural Statistics and sampling through
his excursion to the disciplines of sociology, nutrition and genetics.

19
SCHOOL AND UNDER-GRADUATE STATISTICAL
EDUCATION
I In India a controversy has cropped up in
tertiary statistical education about the curricula and
approaches among those actively associated with
statistical profession. There are wrong beliefs and
notions that statistical understanding is not
necessary to teach statistics. Profession emphasis
on the real problems of statistics; real data and
interdisciplinary interaction; focus on tertiary
teaching; greater awareness of linking with
student’s past, present and future are salient
features of statistical education in India.

The ideas of problem based and collaborative
teaching packages (ref. Densie, 1998) have
influenced statistical teaching.
An approach of structure and real problems in
parallel (ref. Sowey, 1991, 1998) helps students in
developing their concepts and practical
confidence.
Computer technology has expanded horizons for
both learners and teachers accessibility to
techniques, demonstrations and visualization,
enhancing both skill and conceptual
development .
20

21
Realizing the importance of statistics, some
elementary topics were introduced at school level
in the subject of mathematics. The topics included
at 10+1 and 10+2 levels are:
Definition of statistics, Measures of central
tendency, Measure of dispersion, Probability,
Correlation and Regression.

At under graduate level statistics is taught as
one paper in B.A. pass and honors courses in
mathematics in most of the universities, while some
universities have introduced statistics as a separate
subject and the course contents in the curriculum
are:
Descriptive Statistics, Correlation and
Regression, Probability Theory, Random
Sampling, Statistical Estimation, Tests of
Hypothesis and Analysis of Variance.
22

To meet the challenges of teaching statistics to
undergraduate students it is advantageous to adopt the
following :
 Relevant teaching material be prepared and handed over
to the students in the beginning of the course. Textbooks
should be written keeping in view the objectives of the
courses and background of the students and these should be
made available.
The teacher should make efforts to create interest among
the students for statistics learning by correlating its teaching
with real life problems.
 Teaching should be made experimental oriented and
enjoyable illustrating by live examples and using audio-
visual aids 23

24
The above-mentioned objectives can be
fulfilled if prior to teaching a particular statistical
technique, the teacher introduces a situation for
the need of particular technique. This will help
the students to understand the need for the particular
technique and relevance of statistics to their own
discipline. Audio-visual aids like overhead
projector, slide projector and LCD’s can help in
better presentation and condensing information
from bulk and thus provide more information in
short time.

25
Statistics is taught in most of the Indian universities at the
post-graduate level. Universities in some states offer both UG as well
as PG degrees in statistics. ISI Calcutta with its branches in Delhi and
Banglore and Indian Agricultural Statistics Research Institute (IASRI)
at New Delhi are engaged in research and teaching.
In order to impart training at post-graduate level, leading to M.Sc.
and Ph.D. degrees, Calcutta University established the departments of
statistics in 1943. The several other universities also established
separate departments of statistics and started postgraduate program in
statistics. The Universities which are well known nationally and
internationally in statistical education are Calcutta, Delhi, Punjab,
Mathurai, Bombay, Madras, Rajasthan,etc.
HIGHER EDUCATION AND RESEARCH IN STATISTICS

26
In addition to these universities Indian
Statistical Institute at Calcutta, Delhi and
Bangalore are imparting statistical education at
postgraduate level. These institutions and
universities mentioned above deserve the major
share of credit for the past and current developments
of statistical education in India. There is an
International Centre located at Calcutta where
eminent statisticians from every corner of the world
are invited to deliver seminars and talks. In Indian
Statistical Institutes degree courses of statistics are
taught to cover a wide range of subjects somewhat
analogous to courses in medicine and engineering.

27
Most of the universities in India award M.A./ M.Sc. degree
in statistics without thesis, while in some universities a thesis or
dissertation is a partially requirement for postgraduate degree in
statistics. In ISI and IASRI the postgraduate students in statistics are
required to specialize in one of the fields viz. economics, public
health, agriculture, business management, marketing and computer
applications. Experts in their own settings teach them two papers on
the chosen subject. A course ending project work in the real life
setting is a part of curriculum at these universities.
In Agricultural Statistics, the most significant contribution
of Indians is towards the methods of estimation of crops yield
through crop-cutting experiments. In addition to the statistical
wing of the Indian council of agricultural research and the Indian
Statistical Institute, States Agricultural Universities have also
contributed significantly to the development of experimental designs
for judging the optimum size and shape of plots through crop cutting
experiments.

28
The techniques developed by the Indian
Statisticians in areas of Sample Survey, Design of
Experiments, Statistical Genetics, Crop Forecasting,
Categorical Data analysis and Linear Models have
also been adopted by other Asian and African
countries.
Adequate sampling techniques have been
developed for the estimation of area and crop yields.
Methodologies have been developed for the
estimation of fruits and vegetables productions that
are perishable in nature and require multiple
harvesting.

Advanced research has been pursued in both
theoretical and practical aspects of design of
experiments.
In theory, important fields such as asymmetrical
factorial designs, PBIB designs, mating designs
and response surface designs have been
investigated extensively during the last four
decades. Besides construction, other aspects
like optimality and robustness of designs
against loss of observations and presence of
outliers have also been investigated.

30
On applied front, aspects of planning, designing, analysis
and interpretation of data on agricultural crops have been
covered to a large extent.
In recent years much research has been done in statistical
genetics, in particular, applications of genetic statistics to plant
and animal breeding have received attention of many
researchers.
Crop forecasting methodologies have been developed for
obtaining advanced estimates of crop production on the basis
of biometrical characters and weather parameters. In
particular, for crops like rice, wheat, cotton, jute, sugarcane
etc. forecasting models have been developed using various
plant characters measured at different stages of crop growth.

31
(a) The time has come to introduce educational programs
appropriate for statistics as a fully developed new technology that calls
for the utilization of a wide range of scientific knowledge to help in
solving scientific or practical problems. As Fisher had pointed out
“A profession statistician as a technologist must talk the language of
both theoretician and practitioner”. The education of a statistician, like
that of other technologists, must have a broad base.
(b) The challenges in statistics education are never trivial, but offer
opportunities inherent to the richness of statistics as a discipline and a
servant. University structure should tend to encourage competition
rather than cooperation among faculties.
SUGGESTIONS AND RECOMMENDATIONS

32
(c) The awareness in the statistical profession and importance of
statistics education is the need of the hour. The commitment of
teachers to their students and good practice of statistics will lay
excellent foundation for the future. However, the challenge to increase
awareness and acknowledgement across the country needs cooperation
with mathematical sciences and other disciplines.
(d) The future of statistics needs to be discussed in view of recent
development in information technology such as data missing, data
communication and information processing networks and other
requirement of end users.
(e) The current statistical methodology based on probabilistic models
developed for the analysis of small data sets appears to be inadequate
and require some methods to be put forward in the name of data
mining for such purposes.

33
(f) A statistician must have rigorous training on the analysis
packages such as SPSS, SAS etc. During the training period he may
be asked to handle practical problems, case studies and small research
projects of applied nature. Training should also include the
preparation of layout for field experiments and their actual
implementation in the field.
(g) Developments in Statistics are beneficial for both the government
departments and private industry. Therefore, research scholars and
faculty from the universities and research institutions involved in
statistical research must be provided with fellowship and research
grants from government departments and private industries to the tune
of advanced countries.
(h) In the 20th
century, statistics was considered a fundamental
scientific tool, and the basis for the experimental scientific method.
Nowadays it is not possible to think of research projects without
statistical support. Thus it is imperative to include this discipline in
the curriculum at undergraduate level and post-graduate studies.

34
(i) Global communication and an expansion interest in education
are leading to an increasing number of successful international
research or educational programs at different levels: for example,
large scale statistical literacy comparative studies; regional,
national or international funded projects; international statistical
education centers; international training programs or conferences in
statistics education. So it has become imperative to have
international cooperation in statistics education.
(j) There is a need of globalization and diversity in teaching and
research. There are contrasting global and local approaches to
statistics education, however, these approaches usually
complement each other: for example; large sample quantitative
studies versus qualitative ethnographic research; the need to
recognize global tendencies, etc.

35
1. Dansei, B (1998) Using Collaborative Learning Packages to
Teach Introductory Statistics at Post- Dansie Secondary Level.
Pro. 5th
International Conference on Teaching Statistics, 125-
130.
2. Mahalanobis, P.C.(1965) Statistics as a Key Technology. The
American Statistician,43-46.
3. Sowey, E.R. (1991) Teaching Econometrics as Though
Coherence Matters. Pro. of Third International Conference on
Teaching Statistics, 321-331.
4. Sowey, E.R. (1998) Statistical Vistas: Perspectives on Purpose
and Structure. Journal of Statistics Education, vol. 6(2).
REFERENCES

Statistical education

More Related Content

What's hot (20)

Viewers also liked (8)

Similar to Statistical education (20)

More from Department of Mathematics and Statistics, Ramjas College, Delhi University (18)

Recently uploaded (20)

Statistical education