Information Technology Reference
In-Depth Information
−
0.3
5
r
RGF
−
0.2
4
g
3
SLF
−
8
7
q
2
−
0.1
6
−
6
u
5
v
h
1
4
−
4
3
j
c
2
−
2
m
0
p
1
−
1
i
0
f
−
2
−
3
b
n
2
−
1
a
4
0.1
−
2
d
6
k
w
8
s
e
10
0.2
t
12
P
14
16
0.3
SPR
PLF
Figure 2.27
Interpolation biplot of the aircraft data with obliquely translated biplot
axes, such that the point of concurrency is zero on each axis. The original origin is
retained and marked with a black cross. A new point P is interpolated algebraically with
values
SPR
=
8,
RGF
=
4,
PLF
=
0.3 and
SLF
=
3.
By interactively turning on and off the plotting of individual axes, very useful biplot
displays of large data sets can be obtained. Figure 2.28 clearly shows three subsets in
the variables relating to reasons for buying by mail catalogue, perceived risks in buying
by mail catalogue and risk relievers when buying by mail catalogue. The top panel of
Figure 2.30 suggests that
Q49
is out of its place or that perhaps the respondents wrongly
interpret it. Furthermore, the axes relating to reasons for buying by mail catalogue are
closer to one another than the other two sets of axes. This suggests a larger correlation
among questions relating to the reasons than between questions relating to perceived
risks and risk relievers when buying by mail catalogue.
This leaves us with the task of reducing the amount of ink in a biplot with many
sample points. Of course, as in the case of axes, this goal can be accomplished by
interactively turning on and off the sample points. A more fruitful idea is to enclose a
specified subset of sample points and suppress plotting of the samples within the enclosure
while showing only its boundary. We explore several possibilities for enclosing sample
points in a two-dimensional biplot in the next section.
