Agriculture Reference
In-Depth Information
statistical theories towards advancement of
human civilization. Statistics is just like a
molded clay; one can make God or devil out of
it as per the choice of the user. Unfortunately, in
many of the cases, statistical calculations have
been made with the help of the computer
packages, without knowing the logic and
utilities. Misuse of statistical concept is increas-
ing day by day. We must be cautious about
garbage in, garbage out (GIGO). The researcher
must have a clear idea about what are the
requirements, what are inputted to the computer,
and what are the commands to be given to the
computers to get the required information
analyzed properly befitting to the requirement
of the objectives of the research program.
A computer does not have any brain; the only
thing it can do is the use of stored program as per
the direction of the user. Using a set of data, one
can get innumerable types of output just by
changing instruction to the computer. But defi-
nitely all of these outputs are neither relevant nor
correct under the given situation. For best use of
the computer, it is feasible to get from the infor-
mation fed to the computer using statistical
theories, what should be the direction to the
computer and what is the output generated by
the computer. In the following example, we
shall demonstrate how many types of statistical
analysis could be taken up. While using statisti-
cal software, the following points are required to
be noted:
1. The background information about the data
2. Nature of the data
3. Objectives of the experiment
4. Hypothesis to be tested
5. Appropriate statistics to be calculated for 3
and 4 using the specific type of data
6. Background knowledge about the statistical
package to be used w.r.t. its algorithm,
techniques used in performing the task
7. Testing the accuracy of the software using
solved identical example
8. Knowledge about the explanation of the out-
put after running the statistical package and
linking these with the objective of the study
Example 13.1
The following table gives yield
(q/ha) of a particular paddy variety. The experi-
ment was conducted with three types of manure
and three doses of nitrogen to standardize the
supplement of inorganic nitrogen by manure,
and the experiment was conducted under field
condition for two consecutive seasons. The prob-
lem is to find out the best dose of nitrogen and
best manure to get the best yield.
Nitrogen
N1
N1
N1
N2
N2
N2
N3
N3
N3
Rep
R1
R2
R3
R1
R2
R3
R1
R2
R3
Manure 1
Season 1
57.80
65.06
67.78
68.77
70.37
76.47
76.07
64.67
75.37
Season 2
58.43
65.69
68.41
71.37
61.01
68.07
68.81
69.31
70.10
Manure 2
Season 1
76.27
66.37
69.17
59.47
71.47
72.77
69.77
77.63
71.84
Season 2
70.11
66.29
70.14
68.31
74.30
73.72
70.82
71.56
72.87
Manure 3
Season 1
69.07
87.74
81.70
83.07
90.48
72.66
71.86
62.92
61.80
Season 2
73.71
75.38
72.34
73.81
73.12
73.26
72.61
71.76
68.54
Before analyzing the data information on the
type design followed is a must. But most fre-
quently it is found that after completion of the
experiment, the researcher searches for appropri-
ate method of data analysis. Actually in designs
of experiments, once the experimental design is
fixed, its analysis is also fixed. In the absence of
full information, the above data could be
analyzed in various forms, and the results of all
these analyses are not identical. Here, some of
the possible analyses along with experimental
design assumed in each stage are provided.
Readers may note that these are not exhaustive;
one can have other types of analysis with the
same data. We have presented the ANOVA
tables only, from which it is clear that there
exist differences among the analyses and as
such the interpretation will also vary. Thus,
appropriate analysis of the data is required to
extract the information. Otherwise, misleading
