Biology Reference
In-Depth Information
polynomial regressions combined with the four variables with the highest correlations
independent of the 25 regression models employed (including the linear one).
Table 3. Twenty four different regression models applied to detect which could best
estimate the ages of pink river dolphins.
24 different regression models
1- Age = bX
2 - Age = 1/(a + b X)
3 - Age = a + b X + c/X
4 - Age = a + b/X
5 - Age = X/(a + b)
6 - Age = a + b/X + c/X
2
7 - Age = a + b X + c X
2
8 - Age = a X + b X
2
9 - Age = a X
b
10 - Age = a b
X
11 - Age = a b
(1/X)
12 - Age = a X
(bX)
13 - Age = a X
(b/ X)
14 - Age = a e
(bX)
15 - Age = a e
(b/x)
16 - Age = a + b ln X
17 - Age = 1/(a + b ln X)
18 - Age = a b
X
X
c
19 - Age = a b
(1/X)
X
c
20 - Age = a e
(X - b)
(2/C)
21 - Age = a e
(ln X - b)
(2 / C)
22 - Age = a X
b(1- X)
c
23 - Age = a (X /b)
ce
(X /b)
24 - Age = 1/(a (X + b)
2
+ c)
R
ESULTS
Simple Lineal Regression
Out of 50 simple lineal regressions carried out, 24 skull morphometric variables showed
significant relationships with age (p ≤ 0.05) (Table 4). Five of these variables yielded
probabilities of 0.00001. Thus, this indicated a very high correlation with age. These
characters were as follows: Width of the face at the skull base (V3), maximum width through
the zygomatic process (V15), internal longitude of the skull (V21), maximum width of the
internal nares (V30), and width of the left condole (V42). The variable, that explained the
highest proportion of the age, was the maximum width between the zygomatic processes of
Search WWH ::
Custom Search