I would like to understand why an anova is made on a lme (or lmer) and how to interpret the anova output.
lm1 <- lme(EWL..mg.h. ~ Condition*Session + Condition*Sex +
Condition*Mass, random = ~1|Individu,
data = data_respiro_acoustic)
summary(m1)
Anova(m1, type = c("III"))
Summary Output :
Linear mixed-effects model fit by REML
Data: data_respiro_acoustic
AIC BIC logLik
1232.332 1261.459 -606.1662
Random effects:
Formula: ~1 | Individu
(Intercept) Residual
StdDev: 18.01137 13.9361
Fixed effects: EWL..mg.h. ~ Condition * Session + Condition * Sex + Condition * Mass
Value Std.Error DF t-value p-value
(Intercept) 32.45663 25.611020 103 1.267292 0.2079
ConditionS 17.14916 17.503442 103 0.979759 0.3295
SessionSession_2 2.68361 3.284770 103 0.816986 0.4158
SexM 8.13412 7.651439 33 1.063083 0.2955
Mass 0.03087 0.107194 33 0.288001 0.7751
ConditionS:SessionSession_2 -5.78778 4.645366 103 -1.245925 0.2156
ConditionS:SexM 21.74791 5.193698 103 4.187364 0.0001
ConditionS:Mass 0.00369 0.072762 103 0.050762 0.9596
Correlation:
(Intr) CndtnS SssS_2 SexM Mass CS:SS_ CnS:SM
ConditionS -0.342
SessionSession_2 -0.064 0.094
SexM -0.558 0.188 0.000
Mass -0.980 0.330 0.000 0.447
ConditionS:SessionSession_2 0.045 -0.133 -0.707 0.000 0.000
ConditionS:SexM 0.189 -0.554 0.000 -0.339 -0.152 0.000
ConditionS:Mass 0.333 -0.973 0.000 -0.152 -0.339 0.000 0.447
Standardized Within-Group Residuals:
Min Q1 Med Q3 Max
-2.1236561 -0.6115711 0.1679837 0.6287077 2.3180579
Number of Observations: 144
Number of Groups: 36
Anova Output
Analysis of Deviance Table (Type III tests)
Response: EWL..mg.h.
Chisq Df Pr(>Chisq)
(Intercept) 1.6060 1 0.2051
Condition 0.9599 1 0.3272
Session 0.6675 1 0.4139
Sex 1.1301 1 0.2877
Mass 0.0829 1 0.7733
Condition:Session 1.5523 1 0.2128
Condition:Sex 17.5340 1 2.822e-05 ***
Condition:Mass 0.0026 1 0.9595
I would also like to know what type of anova should I do and why.
