Geoscience Reference
In-Depth Information
1.0
.5
P=.1
.2
P=.05
.1
.05
P=.01
.02
P=.005
P=.001
.01
.005
P=.0005
P=.0001
.002
.001
.0005
.0002
.0001
.00005
.00002
.00001
n = Sample Size
Fig. 2.2 Size of random sample (
n
) needed to detect a species occurring with proportional
abundance (
p
) with probability of failure to detect its presence fixed at
P
(After Dennison and
Hay
1967
) (Source: Agterberg
1990
, Fig. 3.1)
2.2.4 Other Discrete Frequency Distributions
Suppose that a succession of Bernoulli trials is started by randomly selecting black
or white cells from a black-and-white mosaic (e.g., an array based on pixels). One
may ask the question of how many times the trial must be repeated before
X
¼
Σ
m
X
i
¼
r
where
r
is a fixed positive integer number and
m
represents the number of
trials it takes to obtain
r
. It is convenient to write
m
k
+
r
where
k
is another integer
number. The probability of a one (for a black cell) at the
m
-th trial is
p
.
This probability must be multiplied by the probability that there were exactly
k
zeros (for white cells) during the preceding
m
-1 (
¼
¼
k+
r
1) experiments.
Search WWH ::
Custom Search