Hi guys, I am trying to run a multi-level analysis using Mplus. My within variables are period (binary X variable) and a probability score (continuous Y variable named PosProb). The between variable is whether the specific score is for a chain hotel or not, so it is also binary. I used hotel ID as the cluster.
Below is the syntax and also the Error message I got. Anyone can give me some advice what I should do? I know that the value for the between variable should be the same for the same cluster ID, and I checked that they are the same, that is for a particular hotel, it is either chain or non-chain. I don't understand why I get the error message.... GREAT THANKS!
variable: names = No ID hotel Rating Chain date reviewID
PosProb NegProb Conf Senti Period;
usevar = Chain PosProb Period;
Cluster=ID;
within = Period;
between = Chain;
ANALYSIS: TYPE = TWOLEVEL RANDOM;
MODEL:
%within%
betalj | PosProb on Period ;
%between%
PosProb on Chain;
betalj on Chain;
PosProb with betalj;
Output: sampstat CINTERVAL;
*** ERROR
One or more between-level variables have variation within a cluster for
one or more clusters. Check your data and format statement.
Between Cluster ID with variation in this variable
Variable (only one cluster ID will be listed)
CHAIN 12
This message usually means that there is a data error such that a Level 2 variable takes on different values for individuals who belong to the same cluster (hotel), which would be impossible. That is, the same hotel could not be both chain and non-chain. I would check again to see if something was entered incorrectly. Another possibility could be that Mplus isn't reading your data file correctly for some reason. If you cannot figure it out, I would contact the Mplus support directly. They should be able to find the problem.
@christian geiser thanks! You are right! I think mplus misread my data because of the formatting issue. I hope there is an easier way to adjust that rather than doing it manually. I will look into it. At least I know it works! Thanks again!!
- Badges
- Science topic
- Similar topics
- Names