One-way random effects model in (2.39) should be used if the treatments under comparison are sampled randomly from a population and the objective is to make inference about the population. The ANOVA...

One-way random effects model in (2.39) should be used if the treatments under

comparison are sampled randomly from a population and the objective is to

make inference about the population. The ANOVA table and the F test are the

same as in the fixed effects case. Additional analysis includes the estimation of

the random effects variance 𝜎2

𝜏 and the population mean 𝜂 (Section 2.5).