Geoscience Reference
In-Depth Information
Table A3
Statistical data for the multiple linear regression OLS models
a
Intercept,
a
Mass coefficient,
b
1
Sample size coefficient,
b
2
Interaction coefficient,
b
3
r
2
Response
Est.
t
p
Est.
t
p
Est.
t
P
Est.
t
p
df
F
Out-degree
b
2.55
3.59
0.0016
0.36
1.94
0.0646
0.05
4.60
0.0001
0.00
0.58
0.5690
23
0.79
29.01
In-degree
c
8.27
6.97
0.0000
1.59
5.67
0.0000
0.04
2.53
0.0189
0.03
2.43
0.0236
23
0.90
65.45
Prey range
3.71
9.61
0.0000
0.01
0.055
0.96
0.01
2.20
0.0500
0.00
0.12
0.91
11
0.61
5.65
a
The regressions were expressed as response variable
¼aþb
1
log
10
(SC mass)
þb
2
SC sample size
þb
3
log
10
(SC mass):SC sample size
þe
, where SC mass:
SC sample size denotes the interaction between the two predictors SC mass and SC sample size,
the unexplained
variance. The sample size of a size class (SC) was the number of individual prey items for the in-degree (vulnerability) and prey-range responses, and the
number of unique predator individuals for the out-degree (generality) response. Bold values indicates a significance level of
a
,
b
1-3
are the regression coefficients and
e
= 95%.
>
b
Vulnerability.
c
Generality.
