The paper evaluates the methods for assessing the significance of fixed effects in linear mixed-effects models using the lme4 package in R. It highlights the limitations of the default methods, such as likelihood ratio tests (LRTs) and the t-as-z approach, which are somewhat anti-conservative, especially for smaller sample sizes. The author conducts simulations to compare these methods with other approaches, including parametric bootstrapping, the Kenward-Roger and Satterthwaite approximations for degrees of freedom. The results show that while all methods have some anti-conservative bias, the Kenward-Roger and Satterthwaite approximations, particularly when applied to restricted maximum likelihood (REML) models, produce more accurate Type 1 error rates and are less sensitive to sample size. Parametric bootstrapping also shows promise but is more sensitive to sample size. The paper recommends using these alternative methods, especially for smaller sample sizes, to ensure more reliable hypothesis testing in mixed-effects models.The paper evaluates the methods for assessing the significance of fixed effects in linear mixed-effects models using the lme4 package in R. It highlights the limitations of the default methods, such as likelihood ratio tests (LRTs) and the t-as-z approach, which are somewhat anti-conservative, especially for smaller sample sizes. The author conducts simulations to compare these methods with other approaches, including parametric bootstrapping, the Kenward-Roger and Satterthwaite approximations for degrees of freedom. The results show that while all methods have some anti-conservative bias, the Kenward-Roger and Satterthwaite approximations, particularly when applied to restricted maximum likelihood (REML) models, produce more accurate Type 1 error rates and are less sensitive to sample size. Parametric bootstrapping also shows promise but is more sensitive to sample size. The paper recommends using these alternative methods, especially for smaller sample sizes, to ensure more reliable hypothesis testing in mixed-effects models.