model1 <- lme(FOD_FODs ~ 0 + Condition_FODs*language_used_FODs_ind, random=~1|subject_FODs, na.action=na.omit, method = "ML", data=indvar_FODs,control = lmeControl(opt = "optim")) > p_value <- shapiro.test(resid(model1))$p.value > round(p_value, 3) [1] 0 > check_singularity(model1) [1] FALSE > ano <- anova(model1) > round(ano, 3) numDF denDF F-value p-value Condition_FODs 4 3262 0.402 0.807 language_used_FODs_ind 4 40 0.008 1.000 Condition_FODs:language_used_FODs_ind 12 3262 0.053 1.000 > library(predictmeans) > eta_squared(anova(model1), partial=TRUE) # Effect Size for ANOVA Parameter | Eta2 (partial) | 95% CI --------------------------------------------------------------------- Condition_FODs | 4.93e-04 | [0.00, 1.00] language_used_FODs_ind | 8.36e-04 | [0.00, 1.00] Condition_FODs:language_used_FODs_ind | 1.95e-04 | [0.00, 1.00] please report these results in APA style

According to the Shapiro-Wilk test, the residual is not normally distributed (p=0). However, the model is not singular (FALSE) and the effect is not significant (p=1).