variance related warning message in path analysis

1,642 views
Skip to first unread message

Fülöp Attila

unread,
Sep 28, 2013, 11:23:39 AM9/28/13
to lav...@googlegroups.com
Dear lavaan users,

I have a path analysis which runs OK, but gives the following warning message:

"Warning message:
In getDataFull(data = data, group = group, group.label = group.label,  :
  lavaan WARNING: some observed variances are (at least) a factor 1000 times larger than others; please rescale"

Why is this message generated? Is this a problem or can I just ignore this message?

Thank you in advance!

Attila


 

Edward Rigdon

unread,
Sep 28, 2013, 12:10:42 PM9/28/13
to lav...@googlegroups.com
Surprising things happen in SEM when observed variables differ substantially in scale.  the warning is giving you good advice.  You can easily rescale by, for example, dividing your larger-scale variables by some power of 10.  Diving the variable by 10 divides its variance by 100.
--Ed Rigdon

Sent from my iPad
--
You received this message because you are subscribed to the Google Groups "lavaan" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lavaan+un...@googlegroups.com.
To post to this group, send email to lav...@googlegroups.com.
Visit this group at http://groups.google.com/group/lavaan.
For more options, visit https://groups.google.com/groups/opt_out.

yrosseel

unread,
Sep 28, 2013, 1:18:37 PM9/28/13
to lav...@googlegroups.com
On 09/28/2013 05:23 PM, F�l�p Attila wrote:
> Dear lavaan users,
>
> I have a path analysis which runs OK, but gives the following warning
> message:
>
> "Warning message:
> In getDataFull(data = data, group = group, group.label = group.label, :
> lavaan WARNING: some observed variances are (at least) a factor 1000
> times larger than others; please rescale"

99% of the time, this warning is completely harmless; but you should
still check your data, eg by typing:

varTable(fit)

to identify the variables with (very) large variances. Rescale the
variable(s), and fit again.

Yves.


Fülöp Attila

unread,
Sep 28, 2013, 1:35:55 PM9/28/13
to lav...@googlegroups.com
Dear Edward and Yves,

Thank you for your answers. I had one variable with very high variance. I rescaled it and performed the analyses again. As you predicted Yves, nothing has changed within the results, but the warning message disappeared. 

Thank you again for your help!

Attila




2013. szeptember 28., szombat 19:18:37 UTC+2 időpontban Yves Rosseel a következőt írta:
Reply all
Reply to author
Forward
0 new messages