[R-sig-ME] How do I know if the lmer model is right?

[Flávio M]

Dear Members,

Could you check if this analysis is right (compile attached)? I am not
familiarized with temporal variables. I am intent to find the equation to
calculate litter amount throught any value of ndvi. Both variables were
coleted monthly during three years.

I did not find a specific model for time series with two variables.
However, as they were collected every month in the same spots lmer should
works fine, right?

Best regards,
Flavio
-------------- next part --------------
library(readxl)
library(afex)
## Loading required package: lme4
## Loading required package: Matrix
## Loading required package: lsmeans
## Loading required package: estimability
## ************
## Welcome to afex. For support visit: http://afex.singmann.science/
## - Functions for ANOVAs: aov_car(), aov_ez(), and aov_4()
## - Methods for calculating p-values with mixed(): 'KR', 'S', 'LRT', and 'PB'
## - 'afex_aov' and 'mixed' objects can be passed to lsmeans() for follow-up tests
## - Get and set global package options with: afex_options()
## - Set orthogonal sum-to-zero contrasts globally: set_sum_contrasts()
## - For example analyses see: browseVignettes("afex")
## ************
## 
## Attaching package: 'afex'
## The following object is masked from 'package:lme4':
## 
##     lmer
data<-read_excel("ndvi3.xls")

data
## # A tibble: 432 x 6
##    month  year stage  plot        ndvi      litter
##    <dbl> <dbl> <chr> <dbl>       <dbl>       <dbl>
##  1     4     1 Early     1  0.30864501 -2.45986132
##  2     5     1 Early     1  0.17898780 -3.22690800
##  3     6     1 Early     1  0.00871899  0.07673787
##  4     7     1 Early     1 -0.20991200 -1.25453724
##  5     8     1 Early     1 -0.21325549 -2.76453008
##  6     9     1 Early     1 -0.21593250 -4.44981737
##  7    10     1 Early     1 -0.30416150 -4.19808639
##  8    11     1 Early     1 -0.10118885 -5.47412419
##  9    12     1 Early     1  0.22365620 -3.27540914
## 10     1     1 Early     1  0.31648629 -2.02710415
## # ... with 422 more rows
summary(data)
##      month            year      stage                plot     
##  Min.   : 1.00   Min.   :1   Length:432         Min.   :1.00  
##  1st Qu.: 3.75   1st Qu.:1   Class :character   1st Qu.:1.75  
##  Median : 6.50   Median :2   Mode  :character   Median :2.50  
##  Mean   : 6.50   Mean   :2                      Mean   :2.50  
##  3rd Qu.: 9.25   3rd Qu.:3                      3rd Qu.:3.25  
##  Max.   :12.00   Max.   :3                      Max.   :4.00  
##       ndvi              litter        
##  Min.   :-0.30416   Min.   :-5.47412  
##  1st Qu.:-0.17107   1st Qu.:-3.87032  
##  Median :-0.01833   Median :-3.00008  
##  Mean   : 0.00433   Mean   :-2.97930  
##  3rd Qu.: 0.18006   3rd Qu.:-2.24116  
##  Max.   : 0.34047   Max.   : 0.07674
m<-mixed(litter~ndvi*stage+(month|plot),data=data)
## Numerical variables NOT centered on 0 (i.e., interpretation of all main effects might be difficult if in interactions): ndvi
## Fitting one lmer() model. [DONE]
## Calculating p-values.
## Note: method with signature 'sparseMatrix#ANY' chosen for function 'kronecker',
##  target signature 'dgCMatrix#ngCMatrix'.
##  "ANY#sparseMatrix" would also be valid
## [DONE]
anova(m)
## Mixed Model Anova Table (Type 3 tests, KR-method)
## 
## Model: litter ~ ndvi * stage + (month | plot)
## Data: data
##            num Df den Df       F    Pr(>F)    
## ndvi            1 422.55 55.0619 6.473e-13 ***
## stage           2 419.25  8.1364 0.0003414 ***
## ndvi:stage      2 419.07 19.5316 7.761e-09 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
summary(m)
## Linear mixed model fit by REML ['merModLmerTest']
## Formula: litter ~ ndvi * stage + (month | plot)
##    Data: data
## 
## REML criterion at convergence: 1348.8
## 
## Scaled residuals: 
##      Min       1Q   Median       3Q      Max 
## -1.72511 -0.65546 -0.07867  0.71961  2.43766 
## 
## Random effects:
##  Groups   Name        Variance Std.Dev. Corr 
##  plot     (Intercept) 2.12254  1.4569        
##           month       0.05024  0.2241   -1.00
##  Residual             1.27952  1.1312        
## Number of obs: 432, groups:  plot, 4
## 
## Fixed effects:
##                        Estimate Std. Error t value
## (Intercept)            -2.78825    0.09599 -29.046
## ndvi                   -0.37417    0.44922  -0.833
## stageIntermediate      -0.57855    0.14344  -4.033
## stageLate              -0.24506    0.13505  -1.815
## ndvi:stageIntermediate -5.53375    0.91168  -6.070
## ndvi:stageLate         -2.00674    0.61924  -3.241
## 
## Correlation of Fixed Effects:
##             (Intr) ndvi   stgInt stagLt ndv:sI
## ndvi        -0.189                            
## stagIntrmdt -0.676  0.162                     
## stageLate   -0.708  0.121  0.476              
## ndv:stgIntr  0.075 -0.394  0.241 -0.055       
## ndvi:stagLt  0.118 -0.627 -0.083 -0.158  0.300