[R] standard error of difference for mixed effects

[ONKELINX, Thierry]

Dear Tony,

A mixed models is not a good idea if you have only two levels for sites and two times two for phase.

The first problem is a mathematical problems. You are estimating variances based on only two and four values. Which is very small, hence you will not get reliable estimates of the variance.

The second problem is more conceptual. Your question indicates that you are interested in the effect of specific levels of your random effects. The philosophy of mixed models is that you only care about the variability due to the random effects, and not in specific effects.

IMHO een simple linear model will do.

lm(log(Salinity) ~ log(flow)*sites/phase)

HTH,

Thierry

----------------------------------------------------------------------------
ir. Thierry Onkelinx
Instituut voor natuur- en bosonderzoek
team Biometrie & Kwaliteitszorg
Gaverstraat 4
9500 Geraardsbergen
Belgium

Research Institute for Nature and Forest
team Biometrics & Quality Assurance
Gaverstraat 4
9500 Geraardsbergen
Belgium

tel. + 32 54/436 185
Thierry.Onkelinx op inbo.be
www.inbo.be

To call in the statistician after the experiment is done may be no more than asking him to perform a post-mortem examination: he may be able to say what the experiment died of.
~ Sir Ronald Aylmer Fisher

The plural of anecdote is not data.
~ Roger Brinner

The combination of some data and an aching desire for an answer does not ensure that a reasonable answer can be extracted from a given body of data.
~ John Tukey
  

> -----Oorspronkelijk bericht-----
> Van: r-help-bounces op r-project.org 
> [mailto:r-help-bounces op r-project.org] Namens Meissner, Tony (DFW)
> Verzonden: donderdag 16 september 2010 2:58
> Aan: r-help op r-project.org
> Onderwerp: [R] standard error of difference for mixed effects
> 
> I have a dataset relating the effects of engineering works on 
> the level of salinity in a river before and after the works.  
> I have modelled this using linear mixed effects models to 
> determine if the significance and level of the response to 
> the works.  I am wanting to calculate the se of difference at 
> several points. I am not sure of how to calculate the 
> difference and which estimate of se I should be using e.g. is 
> it the random effects or the fixed effects estimates and what 
> formula to use e.g. the one for calculating se using linear models?
> 
> To assist here is one of the models I have formulated
> 
>   log(Salinity) ~ log(flow), random=~log(flow)|sites/phase
> 
> where sites is a factor with levels upstream & downstream, 
> and phase is a factor with two levels before and after.
> 
> Any help would be appreciated.
> 
> Tschüß
> Tony Meissner
> Principal Scientist - Monitoring
> Department for Water | Level 3 28 Vaughan Terrace Berri SA 
> 5343 T 8595 2209  | M 0401 124 971 E 
> tony.meissner op sa.gov.au<mailto:tony.meissner op sa.gov.au>
> 
> Mon | Tue | Wed | Thurs | Fri
> 
> www.waterforgood.sa.gov.au<http://www.waterforgood.sa.gov.au/>
>  | www.sa.gov.au<http://www.sa.gov.au/>
> 
> The information in this e-mail may be confidential 
> and/o...{{dropped:13}}
> 
> 

Druk dit bericht a.u.b. niet onnodig af.
Please do not print this message unnecessarily.

Dit bericht en eventuele bijlagen geven enkel de visie van de schrijver weer 
en binden het INBO onder geen enkel beding, zolang dit bericht niet bevestigd is
door een geldig ondertekend document. The views expressed in  this message 
and any annex are purely those of the writer and may not be regarded as stating 
an official position of INBO, as long as the message is not confirmed by a duly 
signed document.