Geoscience Reference
In-Depth Information
For Period I, the fitted model is
Y
=−
0.0625
+
0.4375X and RSS
=
0.6875
1
and for Period II the fitted model is
Y
=
0.4000
+
0.5091X and RSS
=
2.4727
2
Residual sum of squares (RSS u ) = RSS 1 + RSS 2 = 3.1602
Residual sum of squares under null hypothesis (RSS r )
= 6.5565
(6.5565 3.1602)/2
3.1602/(15 )
F
=
=
5.91
It can be inferred that there is structural change in the yield
pattern between drought and normal periods as F 0.05 (2 ,11) = 3.98.
paired t-test for
assessing the
impact
When the two samples are not independent, but the sample
observations are paired together, then this test is applied. The
paired observations are on the same unit or matching units. It
is often used to compare 'before' and 'after' scores in experi-
ments to determine whether significant change has occurred;
for example, to know the impact of climate change on a yield of
perennial crops over years, assuming the rest of the variations
as constant. Let (x i , y i ), i = 1,…,n be the pairs of observations
and let di i = x i − y i . Our aim is to test H 0 : μ 1 = μ 2.
Test statistic
d
s/ n
t
=
d
n
follows t distribution wit h n − 1 d.f., where d /n
=∑ =1
d
and
i
i
n
s
2
=−∑−
=
1
/(n
1)
(d
d)
2
.
d
i
i
1
Cluster and
discriminant
analysis
Cluster analysis is a technique for grouping individuals or
objects into unknown groups. In agriculture, cluster analysis
has been used for diversity analysis, which is the classification
of genotypes into arbitrary groups on the basis of their charac-
teristics. In agro-meteorology, cluster analysis can be used to
analyse historical records of the spatial and temporal variations
in pest/insect populations in order to classify regions on the
 
Search WWH ::




Custom Search