Graphics Programs Reference
In-Depth Information
ssmaller =
1.2097
Now we compare the calculated
F
with the critical
F
. This can be accom-
plished using the function
finv
on a 95% signifi cance level. The function
finv
returns the inverse of the F distribution function with
df1
and
df2
degrees of freedom, at the value of 0.95. Typing
Freal = slarger^2/ssmaller^2
Ftable = finv(0.95,df1,df2)
yields
Freal =
1.0762
Ftable =
1.5400
The
F
calculated from the data is smaller than the critical
F
. We therefore
cannot reject the null hypothesis without another cause. The variances are
identical on a 95% signifi cance level.
We now apply this test to two distributions with very different standard
deviations, 2.0 and 1.2, respectively.
load('organicmatter_five.mat');
Now we compare the calculated
F
with the critical
F
at a 95% signifi cance
level. The critical
F
can be computed using the function
finv
. We again type
s1 = std(corg1);
s2 = std(corg2);
df1 = length(corg1) - 1;
df2 = length(corg2) - 1;
if s1 > s2
slarger = s1;
ssmaller = s2;
else
slarger = s2;
ssmaller = s1;
end
Freal = slarger^2/ssmaller^2
Ftable = finv(0.95,df1,df2)