Multi-level SEM: Zero variance in Clusters/errors

[shruti athavale]

I am conducting a multi-level SEM analysis in the education context with teacher-student clusters. My model runs and gives me errors like the following:

In lav_data_full(data = data, group = group, cluster = cluster,  ... :lavaan WARNING: Level-1 variable “item” has no variance within some clusters. The cluster ids with zero within variance are: ID1, ID2.

We have zero/negative variance in 60 clusters. Is there a way to deal with this issue without having to omit these clusters?

[Terrence Jorgensen]

lavaan WARNING: Level-1 variable “item” has no variance within some clusters. The cluster ids with zero within variance are: ID1, ID2.

This can happen, especially frequently with small clusters (not much opportunity for variability, especially with, e.g., 5-point response scales).  If your model converges, you can probably ignore this warning.  It is just letting you know, in case you accidentally put a Level-2 variable in the Level-1 model (in which case, it would have 0 variance within ALL clusters).

We have zero/negative variance in 60 clusters

How can you have negative variance?  That happens sometimes when estimating residual variances, but your observed variances (sample stats) cannot be negative.

Terrence D. Jorgensen

Assistant Professor, Methods and Statistics

Research Institute for Child Development and Education, the University of Amsterdam

http://www.uva.nl/profile/t.d.jorgensen