[R-sig-ME] Mixed effects model with unbalanced repeated measures across blocks
M West
m.westinbrook at gmail.com
Sun Dec 17 22:01:12 CET 2017
Hello,
I would like to double check that my model is 'OK'.
I am analyzing whether the density of nests differed across four different
forests. Design: I repeatedly sampled transects in the same four forests
throughout the breeding season (May - September). The question here is not
whether the density of nests differed each month (which is what most
examples I've found tend to ask), but I need to make sure to account for
the autocorrelation from repeatedly sampling each forest (i.e., each
transect at each forest).
This design is unbalanced: some sites were only sampled for 2 months,
others were sampled all 5 months and some sites have more transects
compared to others (to account for differences in the length of edges of
the forest). I don't know how to account for this unbalanced design.
The output is included.
Many thanks in advance
~M
model <- with(data, glmer(Number_of_Nests ~ Forest + Week +
(1|Forest:TransectID), family = poisson, data=data))
summary(model)
anova(model, Type = 'II')
Here is the output:
Generalized linear mixed model fit by maximum likelihood (Laplace
Approximation) ['glmerMod']
Family: poisson ( log )
Formula: Number_of_Nests ~ Forest+ Month + (1 | Forest:TransectID)
Data: data
AIC BIC logLik deviance df.resid
555.1 577.1 -267.5 535.1 57
Scaled residuals:
Min 1Q Median 3Q Max
-5.8501 -1.3650 -0.0699 0.9217 4.9935
Random effects:
Groups Name Variance Std.Dev.
Forest:TransectID (Intercept) 0.5939 0.7707
Number of obs: 67, groups: Forest:TransectID, 19
Fixed effects:
Estimate Std. Error z value Pr(>|z|)
(Intercept) 1.64503 0.54558 3.015 0.002568 **
ForestA 1.03360 0.60045 1.721 0.085182 .
ForestB -0.44092 0.72516 -0.608 0.543171
ForestC 1.90743 0.78368 2.434 0.014935 *
MonthMay 0.30640 0.08737 3.507 0.000453 ***
MonthJune -1.11313 0.13132 -8.477 < 2e-16 ***
MonthJuly -2.63123 0.25254 -10.419 < 2e-16 ***
MonthAugust -1.34959 0.17249 -7.824 5.11e-15 ***
MonthSeptember -1.34612 0.72723 -1.851 0.064168 .
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
[[alternative HTML version deleted]]
More information about the R-sig-mixed-models
mailing list