[R-sig-ME] Question about likelihood ratio test and nested models

[David Duffy]

On Mon, 10 Jan 2011, Xiao He wrote:

> Now, I pool level2 and level3 together to form level2_3, hence resulting in
> a new fixed predictor IV_NEW that has only two levels: level1 and level2_3.
> My question then is: is the new model (M2, shown below) considered to be
> nested within the model shown above (M1)? And can I use likelihood ratio
> test to compare M1 and M2?

Sure.  Consider it rather as testing d3=0, in
Intercept + d23 + d3
I d23 d3
1 0   0
1 1   0
1 1   1

-- 
| David Duffy (MBBS PhD)                                         ,-_|\
| email: davidD at qimr.edu.au  ph: INT+61+7+3362-0217 fax: -0101  /     *
| Epidemiology Unit, Queensland Institute of Medical Research   \_,-._/
| 300 Herston Rd, Brisbane, Queensland 4029, Australia  GPG 4D0B994A v