Answer to Question #252254 in Math for dae

Question

Consider a completely randomized randomized design experiment under the fixed effects linear additive model y_ij=μ+t_i+e_ij where μ is the general effect, t_i is the i^th treatment effect and e_ij ^,s are independently normally distributed random variables with common mean zero and variance σ²(i=1,2,......,k) and (j=1,2,.......,n_i). The objective of the experiment is to test the hypothesis H₀ :t₁=t₂=_ _ _=t_k against H₁: all t_i ^,s are not equal.

Obtain the least squares estimates of the effects μ and t_i ^,s and show that they are unbiased