In: Statistics and Probability
1) Define the following statistic tests
a) A log likelihood ratio test
b) Wald's test
2) How is a log likelihood ratio test is constructed to assess theadequacy of a given model
3) Describe features of the covariance correlation matrix
1) A log likelihood ratio test.
Now consider a general family of distributions:
The null hypothesis H0 will state that the parameters belong to
some subspace
of the parameter space
.
Let x1,x2,...,xn be a random sample of size n>1 from a
population with p.d.f f(x,)
where
, the parameter space is the totality of all points that
can assume.
We want to test the null hypothesis :
Against all alternative hypothesis of the type:
The likelihood function ofthe sample observations is given by
(i)
According to the pronciple of maximum likelihood , the
likelihood equation for estimating any parameter
is given by
(ii)
We can obtain the maximum likelihood estimates for the
parameters
as they are allowed to vary over the parameter space
and the subspace
. Substituting these estimates in above equation (i) we obtain
themaximum values of the likelihhod function for variation of the
parameters in
and
respectively. Then the criterion for the likelihood ratio test is
defined as the quotient of these two maxima and is given by
where
and
are the maxima of the likelihood function (i) with respect to
parameters in the region
and
respectively.
ii) Wald's test: The Wald test is a way to find out if explanatory variables in a given model are significant.
Test statistic is
W=
where
is Maximum Likelihood estimator
= Expected fisher information.
NOTE: I have done the first question please repost rest of the questions. Thank you .