Skip to content
Prev 17906 / 20628 Next

Large mixed & crossed-effect model looking at educational spending on crime rates with error messages

Dear James,

Not centring year would require a much stronger random intercept. This
makes the model harder to fit. It will change the random intercept variance
and the correlation between random intercept and random slope. In theory it
shouldn't change the estimates for the fixed effects. But if your model is
unstable / doesn't converge, then how certain can you be on those estimates.

I don't know of any rule of thumb about the overlap in partially crosses
designs. If the model doesn't converge then see how you can simplify it.
E.g. if your model has county_id, then you could drop the state_id as the
county will model any effect at state level (if state level is not in the
model).

I wrote a blogpost on using the same variable both as fixed and as random
effect: https://www.muscardinus.be/2017/08/fixed-and-random/ One of the
examples uses year as a random intercept.

We've used the INLA package (r-inla.org) to model the gamma distribution.

Best regards,



ir. Thierry Onkelinx
Statisticus / Statistician

Vlaamse Overheid / Government of Flanders
INSTITUUT VOOR NATUUR- EN BOSONDERZOEK / RESEARCH INSTITUTE FOR NATURE AND
FOREST
Team Biometrie & Kwaliteitszorg / Team Biometrics & Quality Assurance
thierry.onkelinx at inbo.be
Havenlaan 88 bus 73, 1000 Brussel
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
///////////////////////////////////////////////////////////////////////////////////////////

<https://www.inbo.be>


Op di 10 sep. 2019 om 08:51 schreef Ades, James <jades at ucsd.edu>: