|
|
Chi-square & type= twolevel random |
|
Message/Author |
|
Student 09 posted on Wednesday, April 13, 2011 - 7:58 am
|
|
|
Hi, what's the reason why there is no chi-square statistic available when using type= twolevel random for modelling random slopes? A reviewer has asked me this question & I would be happy for any citable references? |
|
|
The reason is that there is not a single population covariance matrix for testing the model fit. The explanation for this is given on Slide 46 in the most recent Topic 7 course handout on the website. |
|
Back to top |
|
|