Agriculture Reference
In-Depth Information
Slide 8.16b: Step 3 showing the output for the required descriptive statistics using SAS
Along with the values for different measures
of different component characters, graphical
representations can also be made.
From the above-described analysis, using
three different software clearly reveals that
there are limitations, specifications, and methods
in order to get the desired output. But the
user must have a clear idea about the information
being analyzed, requirement, and knowledge
about the statistical method to verify the results
obtained and about the method of handling the
appropriate software. Otherwise, there are few
chances of mis-conclusion, vagueness, superflu-
ousness, or unwarranted/incorrect conclusion.
yield component analysis of any crop, it is found
that yield is an ultimate variable contributed/
influenced/affected by a number of other factors.
If we consider the yield of paddy, then one can
find that the factors like the number of hills per
square meter, number of tillers per hill, number
of effective/panicle-bearing tillers per hill, length
of the panicle, number of grains per panicle, and
test (1,000 grain) weight of grains are influencing
the yield. Variation in one or more of the above-
mentioned factors results in variations of
the yield. Thus, yield may vary because of varia-
tion in the number of hills per square meter or
variations in the number of tillers per hill or so
on. Again, yield may vary because of variation
in the number of hill per square meter and/or in
the number of tiller per hill and other factors.
When we consider variations in one variable
due to variations in any other variable, then it
becomes a bivariate case. On the other hand,
when the variations of more than two variables
are considered at a time, it becomes a multivari-
ate case. The problem of measuring the degree of
association among the variables is considered
through a correlation and regression analysis. In
this section, we shall first study the correlation
8.3
Correlation Analysis
While conducting research works, a researcher
needs to deal with a number of factors/variables
at a time, instead of a single variable/factor.
And all these variables may not be independent
of each other; rather they tend to vary side by
side. Most of the growth/social/economic and
other variables are found to follow the above
characteristics. For example, while dealing with
