2
$\begingroup$

I am trying to construct a LRT to test the hypothesis $H_0: p \ge p_0$ and $H_1: p < p_0$ where $\alpha = .1$ and $p_0 = .6$ and give a critical region.

Attempt:

$\lambda(x) = \frac{"restricted" MLE}{"unrestricted" MLE} = \frac{L(\hat{\theta_0}|X)}{L(\hat{\theta}|X)}$

I am looking for $P(X \in \Re) \le \alpha$ and $P(\lambda(x) \le c) \le \alpha$ where $C \in (0,1)$

$Reject: \{ x: \lambda(x) \le c \}$

My confusion, is with this little information, how can I calculate $\frac{L(\hat{\theta_0}|X)}{L(\hat{\theta}|X)}$.

$\endgroup$
  • 1
    $\begingroup$ Without any kind of a definition for what $p$ even is? $\endgroup$ – Glen_b Dec 14 '13 at 2:15
4
$\begingroup$

The question is underdetermined, so it cannot be answered. As @Glen_b suggests, you need a defintion of $p$, and you also need a probability model i.e. a definition of $L(\theta|X)$.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.