Unformatted text preview: is ˆ0 = −68.3868.
If H1 is true,
2 kj yjk log (θj ) − θj − log (yjk !) l1 = l (θ1 , θ2 ; Y ) =
j =1 k =1 Sum of terms with θ1 plus sum of terms with θ2 .
The MLEs are
ˆ
ˆ
θ1 = 1.423; θ2 = 0.913
The value of l1 at the MLEs is ˆ1 = −67.0230
l
UNM Notice that regardless of the data ˆ1 ≥ ˆ0 . Why?
l
l
How can we decide if difference in loglikelihood is
signiﬁcant?
Need to know the sampling distribution of l (θ; Y ).
What is the distribution of ˆ1 − ˆ0 ?
I
I
Could also rely on the...
View
Full Document
 Fall '13
 GabrielHuerta
 Trigraph, Maximum likelihood, Errors and residuals in statistics, Akaike information criterion

Click to edit the document details