0 Views
Distance Education Unit, Directorate of Extension, UAS, Hebbal, Bangalore-560024, Karnataka.
Biotechnology is emerging as a millennium tool in Agriculture. The trends for adoption of emerged scientific revolutions
in biotechnology needs to be focused for effective development, dissemination and utilization of technologies pertaining to
biotechnology. Market feasibility, market stability, cost, profit consistency, compatibility, visibility, trial ability and demonstrability
have to be considered while developing the appropriate technologies. Capacity building, skill development, effective
communication skills, training, human resource development, effective programme planning needs to be strengthened among
the biotechnology professionals by utilizing the seven orchestrated, concerted comprehensive Narayana’s wheel model.
KEYWORDS: New plant breeding techniques, biotechnology adoption, Narayana’s wheel model.
The nineteenth century as an era of industrial revolution, 20th century is golden age of electronics and 21st century emerged as biotechnological revolution. However, five Scientific Revolutions have emerged in the world during this millennium like (1) Genetic Engineering and ability to use this understanding to develop new process and products in biotechnology. (2) Ecotechnology and blending of best traditional knowledge with frontier technologies in biotechnology. (3) Information technology and its rapid growth in the systematic assimilation and timely dissemination to the concerned professional. (4) Motivation techniques for efficient utilization of available biotechnology innovations by ultimate users. (5) Need for appropriate policies to biotechnology development, technology dissemination and technology utilization. Hence, it is desirable to know
the details of trends for adoption of emerged scientific
revolutions in biotechnology.
(1) Appropriate Technology Development Through
New Plant Breeding Techniques As A Trend
Techniques for introduction of novel traits more
precisely in plants often without introduction of foreign
genetic materials known as new plant breeding techniques
is also not properly addressed by the biotechnology plant
scientists. However, the several research and evaluation
studies conducted in India revealed that utilization of new
plant breeding techniques in plant science is far from
satisfactory in different parts of the country. However,
Narayana’s Innovation Attributes Lotus Model (Fig.1)
explains four attributes with seven sub components under
each for appropriateness of new plant breeding techniques
in plant science to adopt like Relative advantage,
Compatibility, Practibility and Complexity for livelihood
security in the emerging global order. (A) Relative
Advantage: Is the degree to which an innovation of new
plant breeding techniques is superior to the idea it
supersedes. It can be explained with seven sub items like.
(1) Market feasibility: as the extent of market demand
for the product derived out of new plant breeding
techniques innovation and also the extent of scope for
marketing product. (2) Market stability: as the consistency
of market price and demand of product derived out of
new plant breeding techniques Innovation. (3) Cost: is of
two types, initial cost and continuing cost. Initial cost
represents the capital investment required for adoption
of new plant breeding techniques innovation. Further, the
cash or inputs required for subsequent years use of new
plant breeding techniques innovation is termed as
continuing cost. (4) Net Profit: as the quantum of monetary
benefit obtained by an individual through adoption of new
plant breeding techniques innovation. (5) Profit
consistency: denotes the regularity of net returns obtained
by an individual or group of individuals over a period of
time by adoption of a new plant breeding techniques
innovation. (6) Utility potential: as the degree to which
the multiple use potential of a new plant breeding
techniques innovation to an individual or group of
individuals through adoption of new plant breeding
techniques innovations. (7) Time saving: Indicates the best
efficiency of a new plant breeding techniques innovation
in terms of saving time in different aspects. (B)
Compatibility: a new plant breeding techniques innovation
is consistent with past experiences, existing values, and
future plans of the adopters of plant science technologies.
Compatibility is divided into seven sub groups like (1)
Cultural compatibility: a new plant breeding techniques
innovation is consistent with the values and norms of the
society. (2) Social compatibility: denotes prestige gain or
esteem by individual in the society through adoption of a
new plant breeding techniques innovation. (3) Physical
compatibility: a new plant breeding techniques innovation
is consistent and should fit into the needs and interests of
the adopters. (4) Psychological compatibility: New plant
breeding techniques innovation usefulness as perceived
by the members of social system. (5) Situational
compatibility: denotes consistency and harmony of the
new plant breeding techniques innovation with previous
practices followed by adopters. (6) Relational
compatibility: a new plant breeding techniques innovation
can be adopted independently by the adopters just like
other practices. (7) Anticipated compatibility: A new plant
breeding techniques innovation should be consistent with
the future ideas of the adopter over a period of time. (C)
Practibility: a new plant breeding techniques innovation
of plant science can be easily communicated, tested,
demonstrated and practiced. (1) Point of origin: Indicates
the credibility of the source from where the new plant
breeding techniques innovation originated. (2) Access to
advice: For implementation of new plant breeding
techniques innovation, its extent of availability of original
and detailed information for guidance and clearing doubts
that arise while implementing it. (3) Visibility: the results
of a new plant breeding techniques innovation are visible.
(4) Trialability: The degree to which new plant breeding
techniques innovation can be tried on a small scale. (5)
Mastery: The practice of a new plant breeding techniques
innovation could be learned or mastered in a short period
of time. (6) Demonstrability: a new plant breeding
techniques innovation can be demonstrated to members
of social system easily. (7) Communicability: The
information about the new plant breeding techniques
innovation can be diffused to members of the social system
easily and speedily. (D) Complexity: a new plant breeding
techniques innovation of plant science is relatively
difficult to understand and use. (1) Failure probability: a
new plant breeding techniques innovation chances of
failure and uncertainity of results after its adoption. (2)
Discomfort saving: Represents avoidance of physical
discomfort may be derived by adoption of a new plant
breeding techniques innovation. (3) Resource complexity:
difficulty in getting the necessary inputs and other
resources for the application of a new plant breeding
techniques innovation. (4) Reversibility: degree of ease
with which the new plant breeding techniques innovation
can be replaced in case of its failure. (5) Work efficiency:
the adoption of new plant breeding techniques innovation
saves labour or increase the available labour efficiency.
(6) Cognitive complexity: an extent of relative difficulty
in understanding a new plant breeding techniques
innovation. (7) Application complexity: relative difficulty
of a new plant breeding techniques innovations use and
application on the farm.
(2) Capacity Building among Biotechnology
Professionals as a Trend:
Capacity building among biotechnology plant
scientists is too complex phenomenon to be explained by
a single factor. However, Narayana’s Wheel Model (Fig.2)
explains combination of seven components for capacity
building among biotechnology plant scientists viz.
Innovativeness, Decision making ability, Achievement
motivation, Information seeking ability, Risk taking
ability, Coordinating ability and Leadership ability. The
combined contribution of the above seven factors to an
individual behavior is being expressed in terms of capacity
building among biotechnology professionals, so far
attention given is limited.
(1) Innovativeness: Considered as socio-psychological
orientation of a biotechnology plant scientists closely
associated with change, adopting new ideas and practices.
An individual biotechnology plant scientist adopts new
ideas relatively earlier than others in his/her organization.
However, innovativeness in professionals is very essential
to motivate others for adoption of new plant breeding
techniques in plant science. (2) Decision making ability:
Considered as the nature of decision making either
individually or consulting with others while performing
new plant breeding techniques in plant science activities.
It is the degree to which an individual justifies his selection
of most efficient means from among the available
alternatives on the basis of scientific criteria for achieving
maximum profits. Hence, decision making ability is very
important among biotechnology plant scientists to
motivate professional for adoption of new plant breeding
techniques in plant science. (3) Achievement motivation:
Emerging trends in new plant breeding techniques
Every biotechnology plant scientist has a desire to achieve
certain things in life. Achievement motivation is
considered as the extent to which an individual is oriented
towards maximizing profits. Achievement motivation as
a social value that emphasizes a desire for excellence in
order for biotechnology plant scientists to attain a sense
of personal accomplishment. So achievement motivation
increases efficiency of biotechnology professionals in use
of new plant breeding techniques in plant science. (4)
Information seeking ability: It refers to the frequency of
contact by biotechnology plant scientist with various
information sources. This is the pattern by which a
biotechnology plant scientist gets his/ her information
either on his/her own seeking or as a consequence of his/
her being a part of the network. This component is
important for use of new plant breeding techniques in
plant science by professionals. (5) Risk taking ability:
Some biotechnology plant scientists take more risk, some
others take moderate risk and many biotechnology plant
scientists hesitate to take risk. Risk taking ability
considered as individual orientation towards risk and
uncertainity in adopting new ideas and courage to face
the problems. Use of new plant breeding techniques in
plant science demands ability to take risk by professionals.
(6) Coordinating ability: In order to complete the required
work in stipulated period, a biotechnology plant scientist
has to harmonize and synchronize the various activities
for better profit. It is an individual co-ordinates action in
a time dimension. This ability helps to increase the
efficiency of professional in motivating biotechnology
plant scientists for adoption of new plant breeding
techniques in plant science. (7) Leadership ability: To get
things done properly, a plant scientist has to initiate the
action, motivate the followers and decision should be
taken. It is an individual initiates or motivates the action
of the other fellows. Hence, leadership ability is an
important component in biotechnology professional to
motivate plant scientists for adoption of new plant
breeding techniques in plant science.
(3) Skill Development among Biotechnology
Professionals as a Trend
To meet the requirement of emerged scientific
revolutions the biotechnology plant scientists concerned
should be trained properly. To train biotechnology plant
scientists specific skills are required for conducting an
effective training. The skill as ability to do things, to
effectively apply knowledge and personal aptitudes and
attitudes in work situation. However, the concept of skill
concerns the ability to use ones knowledge effectively
and rapidly in execution of performance and more
generally it is an acquired power of doing something
competently. Further, skills notably professional skills are
becoming increasingly important. Biotechnology today
calls for professional skills in its application. Seven skills
are identified for effectiveness among biotechnology plant
scientists. However, Narayana’s Wheel Model (Fig.3)
describes seven orchestrated, concerted comprehensible
skills required among biotechnology plant scientists for
efficiency like (1) Technical Skill is the ability of the
biotechnology plant scientist to use any technique or
method or equipment or product or process as a tool in
the context of new plant breeding techniques in plant
science. (2) Human Skill is the ability of biotechnology
plant scientists in motivating other professional involved
in new plant breeding techniques in plant science with
thorough understanding while working with them as a
team. (3) Conceptual Skill is the ability of biotechnology
plant scientists coordinating and integrating all the
activities of new plant breeding techniques with visionary
outlook. (4) Managerial Skill is the ability of
biotechnology plant scientists in planning, organizing,
directing, leading, reporting and budgeting and reviewing
the work of other professional involved in new plant
breeding techniques in plant science. (5) Design Skill is
the ability of biotechnology plant scientists in finding out
a workable solution to problems through new plant
breeding techniques requires deliberate efforts to develop
solution. (6) Creative Skill is the ability of biotechnology
plant scientists in generating new ideas or in doing things
already done in a new way through new plant breeding
techniques in plant science. (7) Communicative Skill is
the ability of the biotechnology plant scientists to adopt
technologies at different levels using series of new plant
breeding techniques over a period of time.
(4) Effective Communication among Biotechnology
Professionals as A Trend
Good communication does not consist merely of
giving orders but of creating understanding. It does not
consist merely of imparting of knowledge but to help
biotechnology plant scientists to gain a clear view of the
meaning of knowledge. It is therefore, the responsibility
of professionals involved in generation of new plant
breeding techniques in plant science to familiarize
themselves to become effective communicators. However,
Narayana’s Wheel Model (Fig.4) explains seven
orchestrated concerted comprehensible in the system for
effective communication. (1) Credibility: Climate of
belief, earnest desire of new plant breeding techniques in
plant science. (2) Context: Realities of the situation must
provide for participation and playback for new plant
breeding techniques in plant science. (3) Content: The
new plant breeding techniques message must have
meaning for the biotechnology plant Scientist and the
content determines the plant scientist. (Vice-versa). (4)
Clarity: The new plant breeding techniques message must
be put in simple term, words must mean the same thing to
the biotechnology plant scientist as they do to the other
professional. (5) Channels: Established channels of
communication which biotechnology plant scientists
respects must be used for promotion of new plant breeding
techniques in plant science. (6) Consistency:
Communication is an unending process for biotechnology
plant scientists. However, it requires repetition of new
plant breeding techniques in plant science to achieve
penetration, it should be consistent. (7) Capability: This
refers to availability, habit, ability and Knowledge of
biotechnology plant scientist. Much misunderstanding
results from faulty communication. Too many
biotechnology plant scientists saying the wrong things at
the wrong time, in the wrong ways, to the wrong
professional slows progress. What is needed is more
biotechnology plant scientists saying right things, at the
right time, in the right ways to the right professional. This
is a formula for good and effective communication. The
new plant breeding techniques in plant science promoter
is actually a motivator, needs devotion and full
identification with the biotechnology plant scientists,
which are pre-requisites for success.
(5) Training of Biotechnology Professional as a Trend
Training is a planned and systematic effort to increase
biotechnology professional competency. Further, to enable
the biotechnology plant scientists to increase knowledge,
to improve skills, to inculcate appropriate attitude and
develop appropriate attributes to serve better. Several
training models are used by the organizations to influence
the biotechnology professional to make desirable changes
in their behaviour to achieve the objectives of the
organization. Further, observed that training is a building
process, to reflect this, a good course is organized in
ascending order of complexity. However, understanding
of modern biotechnology and deliver it to users in a
usuable form along with monitoring of activities needed
to implement and evaluate its usefulness are urgently
needed. The information has to be integrated with
available communication methods to suit the resource
positions of institutions and time following the integration.
Further, good linkages have to be established with inter
and intra system of biotechnology plant scientists in
organizations. However, experience gained in training so
far indicates that mere development of conceptual
understanding and an operational plan based on it may
not be adequate. Training has to be made to work. This
can happen only when all the three parties involved in
training like organization, trainer and trainee – join in
their effort and make it to work. An important issue facing
us is commitment to training. This is required and it is
seldom well realized. Hence, there is a need for knowledge
of training models to train biotechnology professionals
in plant science. The Narayana’s Model of training process
(Fig.5) for training biotechnology plant scientists, explains
training process may be a temporary system but the trainer
and trainee both learn through various opportunities
available for checking their effectiveness. This also
explains training as an interdependent and interrelated
process. Here lot of opportunity is given for independent
and intervening variables to become dependent variables.
Hence, this model helps to increase the efficiency of
biotechnology professionals in plant science and to
develop competency among biotechnology plant
scientists. The process of training must start by
questioning the basic assumption which has governed our
training approach. So an analysis of SWOT i.e. , Strength
Weakness and Other Things called for to enable as to have
new conceptualization. The training of biotechnology
plant scientists during the millennium must take in to
account the needs of broad based plant science to introduce
greater professional competence.
(6) Human Resource Development among Biotechnology
Professional as a Trend
During the last five decades of development, growth
of developing countries is directly related to their human
resource bases. The countries which have given good
performance are the countries which have made
significant investments in Human Resource Development
(HRD) of biotechnology professional. There is an over
whelming evidence that human capital is one of the key
factors for adoption of new plant breeding techniques in
plant science in Developed countries. Further, HRD is
widely regarded as the single most important resource
for faster adoption being attempted in the developing
countries. Hence, there is need for training to develop
human resource at various levels in biotechnology
institutions for efficient use of new plant breeding
techniques in plant science. Hence, Human resources are
assuming increasing significance during the millennium.
However, experience in the past has indicated that HRD
among biotechnology professional is lacking in our
country as revealed by large number of research and
evaluation studies. How it should be done is explained in
Narayana’s Algebraic Model of HRD (Fig. 6) among
professionals. HRD = HRS + HRT + HRU i.e., HRD =
HR (S+T+U) i.e., Selection of Human Resource, Training
of Human Resource and using of Human Resource
profitably are urgently needed for adoption of emerging
new plant breeding techniques. This is a challenge that
needs to be tackled immediately.
(7) Effective Programme Planning For New Plant
Breeding Techniques in Biotechnology as a Trend
Biotechnology plant scientists should have effective
programme planning and execution for adoption of
emerging new plant breeding techniques in plant science.
However, Narayana’s Wheel Model (Fig.7) explains seven
important steps in programme planning through (1)
Analyzing the situation, (2) Identifying problems, (3)
Finding solutions, (4) Deciding objectives, (5) Plan of
work, (6) Execution of plan and (7) Evaluate the
NARAYANA’S WHEEL MODEL FOR
Fig. 2. Seven Orchestrated, Concerted, Comprehensible
shown on the Wheel
NARAYANA’S WHEEL MODEL FOR
SKILL DEVELOPMENT
Fig. 3. Seven Orchestrated, Concerted, Comprehensible
shown on the Wheel
NARAYANA’S WHEEL MODEL FOR EFFECTIVE
COMMUNICATION
Fig. 4. Seven Orchestrated, Concerted, Comprehensible
shown on the Wheel
effectiveness of new plant breeding techniques in plant
science programmes concurrently at the end of a year.
The strong and weak points identified may be considered
in revising the subsequent programmes to promote new
plant breeding techniques in plant science. The evaluation
report prepared must reach large number of biotechnology
plant scientists and organizations in the locality and in
similar outside locations. However, the experience in the
past has revealed that for effectiveness it should answer
to local plant scientist needs. To achieve this aim a process
of extensive consultation with the concerned target group
of biotechnology plant scientists is required.
CONCLUSION
A dynamic trend needs to be provided to cater the
needs of different categories of biotechnology plant
scientists. The country needs to be mapped out for
immediate growth potential areas in new plant breeding
techniques and future growth potential areas in
biotechnologies. Governmental system need to be
reoriented with proper POSDCORB (Planning,
Organizing, Staffing, Directing, Coordinating, Reporting
And Budgeting) individual biotechnology plant scientist
need to be educated to use the available technology in a
planned way for maximization of profit.
STATISTICAL DESIGNS AND REGRESSION ANALYSIS IN AGRICULTURAL SCIENCES
G. MOHAN NAIDU* AND P. SUMATHI
Department of Statistics and Maths, S.V. Agricultural College, Tirupati – 517 502
ABSTRACT
Statistics is important in the field of agriculture, because it provides tools to analyze collected data. Many modern statistical
techniques were first developed for use in agricultural research, and many basic statistical tools are still important for such
research. Good experimental design, following the basic principles of experimental designs, allows the control of anticipated
environmental variation and the estimation of treatment effects in the presence of such variation. ANVOA provides a wideranging
approach to the analysis of data from designed experiments, aiding the interpretation of the results of complex experiments.
Regression analysis can be used to explore the relationships between a quantitative response variable and one or more quantitative
explanatory variables.
KEYWORDS: ANOVA, Experimental design, GLM, Hypothesis testing, Regression, p-value, Variability
INTRODUCTION
Statistical education for agriculturists tries to give
them a solid foundation in statistics. A wide use of
statistical methods in order to allow the students to apply
these techniques in many fields of agricultural sciences
like filed crop production, livestock, veterinary medicine,
agricultural mechanization, water resources, agricultural
economics and other fields. The use of statistical
techniques in agriculture goes back many years and in
fact, many of the modern statistical techniques were first
developed for use in agricultural research. Early
developments, due to R.A. Fisher at Rothamsted
Experimental Station in the United Kingdom in 1920s
included the basic principles of experimental design –
replication, randomization and local control – and the
analysis of variance (ANOVA), and these techniques, in
common with many statistical methods, were developed
to cope with the inherent variability associated with
experimentation using biological material. In fact, it is
the need to explain or allow for the extensive variation
often found in experimental biological data that has
driven, and still drives, the development of statistical
techniques. By using the correct statistical tools, we can
separate the signal from the noise within our data – if we
do not handle the experimental variability properly we
run the danger of being unable to draw any useful
conclusions from our data.
DESIGN OF EXPERIMENTS AND ANALYSIS
OF VARIANCE
In the design of experiments, the grouping or
blocking of experimental units can be used to eliminate
the effects of systematic changes in environmental
conditions (the experimental units within a block are
assumed to be as similar as possible). The randomization
of treatments to units can protect against unknown
variability, replication provides the basis for the
comparison of treatments, allowing the assessment of
whether the differences between treatments are large
relative to the variation between replicate observations
on each treatment. The most commonly used experimental
design is the “randomized complete block design”, with
a complete replicate of the set of treatments appearing in
each block of experimental conditions. These include
incomplete block designs, row-and-column designs (e.g.
Latin squares) and spilt plot designs. The analysis of
variance technique separates the variation in observed
results into that due to the applied treatments and that
due to the experimental environment, and hence allows
the assessment of whether observed treatment differences
are important relative to the underlying variation between
experimental units. The ANOVA technique for analyzing
data from designed experiments is readily available in
most statistical computing packages.
REGRESSION ANALYSIS
Where applied treatments are quantitative, it is often
of more interest to determine the form of relationship
between the response variable and these explanatory
variables using regression analysis. Simple linear
regression is concerned with fitting the simplest of
relationships, a straight line, between the response variable
and a single explanatory variable, with the parameters of
the line (slope, intercept) determined to minimize the
variance in the response variable about the fitted line. It
is important to realize that the adjective linear in simple
linear regression refers not to the fitting of a straight line,
but to the relationship between the response variable and
parameters being linear.
Extensions of this linear regression approach include
multiple linear regression (more than one explanatory
variable), linear regression with groups (including a
qualitative treatment factor and allowing parameters to
vary with different levels of this factor) and polynomial
regression (quadratic, cubic… etc., relationships between
response variable and explanatory variables). Many real
biological relationships, however, are not well described
by the range of models that can be constructed within the
linear regression framework, but require the use of models
where the response variable is related to the parameters
in a non-linear fashion. Advances in computer power now
make the fitting of such non-linear regression models
relatively simple, and many standard non-linear response
functions are readily available in most statistical
computing packages. These include models based on the
exponential function (for example, to describe the decay
of pesticides in soil or unconstrained growth), sigmoid
functions, such as the logistic and Gompertz curves (to
describe constrained growth or for dose-response studies),
and rational functions, including inverse polynomials
(used to describe the relationship between crop yield and
applied nutrient levels).
The ANOVA and regression analysis methods that
are mentioned above have an underlying assumption that
the response variable is continuous and normally
distributed. However, much of the data collected in
agricultural research, particularly in relation to crop
protection research, are in the forms of discrete counts
(numbers of weeds, insects, disease lesions) or proportions
based on counts (numbers of diseased fruit per tree, or of
insects killed by some treatment), and therefore do not
satisfy these assumptions. For example, count data may
follow a Poisson distribution and proportions based on
counts may follow a Binomial distribution. In this situation
two approaches are possible – to find some transformation
of the data that allows this assumption to be satisfied or
to use an alternative form of analysis that takes account
of the distributional form of data. The development of
General Linear Models (GLMs) by McCullagh provided
a solution to the latter approach, allowing the analysis of
data for a range of non-normal distributions, within the
same basic structure as for analysis of variance and
regression analysis. Of particular interest within this frame
work are log-linear models for count data and probit/logit
models for proportions based on counts, these latter
approaches being particularly important for the analysis
of bioassay experiments.
There are a number of areas where future
development of statistical methodology will be important
in agriculture. One is the analysis of spatial data. Whilst
spatial statistical methods have been developed and used
for many years, particularly geo-statistical methods in the
mining industry and hydrology, there has been relatively
little use of such methods in agriculture. Interest in the
spatial distributions of plants, pests, diseases, nutrients,
and pesticides, however, is now becoming important both
in understanding the biological processes behind
agricultural production and particularly in the
development of precision agriculture approaches to apply,
for example, pesticides or fertilizers to match the
requirements of small areas of crop. Another area where
development of statistical methodology is needed is for
on farm experimentation, involving the assessment of
experimental methods when scaled-up from small
experimental plots to whole field (or even whole farm)
experiments.
STATISTICAL ERROR IN HYPOTHESIS
TESTING
There are two types of error or incorrect conclusions
possible in hypothesis testing and possibilities in which
the statistical test falsely indicates that significant
differences exists between the two or more groups and
also analogously to a wrong positive results. Rejection of
null hypothesis (H0) when it is true is called Type-I error
and acceptance of null hypothesis (H0) when it is false
and it is known as Type-II error and Type-II error is more
harmful than Type-I error (Keppel, 1978; Gupta and
Kapoor, 1970).
The probability of Type-I error is known as level of
significance (á) and the probability of type II error is
known as the power of the test â or (1-á) (Keppel, 1978;
Gupta and Kapoor, 1970). By convention, statistical
significance is generally accepted if the probability of
making Type-I error is less than 0.05, which is commonly
denoted as p<0.05 (Elenbaas et al., 1983). The probability
of Type-II error is more difficulty to derive than
probability of type-I error, actually it is not one single
probability value. The probability of type-II error (â) is
often ignored by researcher (Freeman et al., 1978). The
probability of type- I error (á) and probability of type-II
error (â) are inter-related. As á arbitrarily decreased, â is
increased. Similarly, á is increased, â is decreased
(Hopkins and Glass, 1978; Keppel, 1978).
P-VALUE
The p value is the probability to observe effects as
big as those seen in the study if there is really no difference
between the groups or treatments. The reasoning of
hypothesis testing and p values is convoluted. The p values
helps to answering whether this apparent effect is likely
to be actual or could just by chance or sampling
fluctuation. The p values give the magnitude of difference
present between populations. In calculation of p values,
first assume that no true difference between the two
groups/treatments. The p values allow the assessment of
findings that are significantly different or not. If the p
value is small, the findings are unlikely to have arisen by
chance or sampling fluctuations, reject the null hypothesis.
If the p is large, the observed difference is plausibly chance
finding, we do not reject the null hypothesis. By
convention, p value of less than 5 per cent is considered
small or significant. Sometimes p value is less than 1 per
cent or 0.01, called as highly significant (Gupta and
Kapoor, 1970; Rao, 1985).
CONCLUSIONS
Many modern statistical techniques were first
developed for the use in agricultural research, and many
basic statistical tools are still important for such research.
Good experimental design, following the basic principles
of replication, randomization and local control, allows the
control of anticipated environmental variation and the
estimation of treatment effects in the presence of such
variation. ANVOA provides a wide-ranging approach to
the analysis of data from designed experiments, aiding the
interpretation of the results of complex experiments.
Regression analysis can be used to explore the relationships
between a quantitative response variable and one or more
quantitative explanatory variables. Linear regression
techniques primarily provide an explanatory approach,
whilst non-linear regression techniques allow the modeling
of responses using biologically realistic relationships.
Generalized linear models (GLM) provide an important
tool for working with the non-normally distributed data
that is common in the crop protection experimentation that
frequently occurs in agricultural research, with log-linear
models (for count data) and probit or logit models (for
counts as proportions) being important specific cases.
Future developments of statistical methodology will be
important in three areas of agricultural research – the
analysis of spatial data, the development of precision
agriculture techniques, and on-farm experimentation. In this
paper, the role of statistical research design and regression
application of basic techniques in agricultural research, have
been emphasized scientifically.