Cutpoints in Monte Carlo for cross-cl... PreviousNext
Mplus Discussion > Multilevel Data/Complex Sample >
 Yaacov Petscher posted on Monday, May 02, 2016 - 11:02 am
I am attempting to generate a cross-classified IRT simulation with a binary person-level indicator. I attempted to use "cutpoint" but finding that coupled with my statement on the between level for individuals leads to an error of, "Unrestricted x-variables for analysis with TYPE=CROSSCLASSIFIED must be a variable on only one level". In my code below I only have it specified on one between-level. I greatly appreciate any thoughts on restructuring the code.
names are u x; cutpoints=x(0);
nobservations = 6480;
nreps = 1;
CSIZES = 36[180(1)];
ncsizes = 1[1];
ANALYSIS: TYPE = cross random;
estimator=bayes; proc=2;
model population:
%between level2a%
f BY u@1;
f on x*.5;

%between level2b%
u@1; [u$1*0];
%between level2a%
f BY u@1;
f on x*.5;

%between level2b%
u*1; [u$1*0];
 Tihomir Asparouhov posted on Wednesday, May 04, 2016 - 10:17 am
Insert this line after "categorical=u;"

between=(level2a) x;
 WEN Congcong posted on Sunday, January 01, 2017 - 2:53 am
Dear professors,

Hello! I have one question about the cut points. In example 12.1, the syntax defines the cutpoints = x2(1) and explains that ¡°the variable x2 is cut at the value of one which is one standard deviation above the mean because the mean and variance used for data generation are zero and one. This implies that after the cut x2 is a 0/1 binary variable where 16 percent of the population have the value of 1.¡±

Because the variance is 1, the value 1 is one standard deviation. But I don¡¯t understand the last sentence. (1) How can a 0/1 binary variable have a mean of 0? And how do you know 16 percent of the population have the value of 1?

(2) what about the specification cutpoints = x(1) w(0); in example 12.6 ? How do we comprehend the w(0)?

Thank you very much!
 Linda K. Muthen posted on Sunday, January 01, 2017 - 7:33 am
The original variable before being cut has a mean of zero and a standard deviation of one. The percentages are taken from a z-table.

The cutpoint of zero implies a fifty/fifty split.
 WEN Congcong posted on Sunday, January 01, 2017 - 4:50 pm
Thanks a lot!
Back to top
Add Your Message Here
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Options: Enable HTML code in message
Automatically activate URLs in message