I am helping a colleague with stats analysis, and though it's a seemingly
simple setup, it's becoming quite complicated!
The system is a deciduous forest with treefall gaps of different carefully
chosen sizes. The response variable is amount of NH3 found in the rainwater
collected under each
On 11/29/2010 11:05 AM, Mudrak, Erika [EEOBS] wrote:
I am helping a colleague with stats analysis, and though it's a
seemingly simple setup, it's becoming quite complicated! The system
is a deciduous forest with treefall gaps of different carefully
chosen sizes. The response variable is
One could also setup an SEM with autoregressive paths specified. The missing
data issue is something separate, however. There are ways in SEM of dealing
with it - Full Information Maximum Likelihood and the like. I'd recommend
taking a look at the lavaan package in R.
Of Ben Bolker
Sent: Tuesday, 30 November 2010 12:23 AM
To: r-sig-ecology@r-project.org
Subject: Re: [R-sig-eco] SEM with time series?
On 11/29/2010 11:05 AM, Mudrak, Erika [EEOBS] wrote:
I am helping a colleague with stats analysis, and though it's a
seemingly simple setup, it's becoming quite
