Now suppose you are told that the true demand function is

$ln{Y}_t={}_1+{}_{2}ln{x}_{2t}+{}_{3}ln{x}_{3t}+{}_{6}ln{x}_{6t}+u_t$

but you think differently and estimate the following demand function:

$ln{Y}_t={}_1+{}_{2}ln{x}_{2t}+{}_{3}ln{x}_{3t}+u_t$

a$)$ Carry out RESET and LM tests of specification errors, assuming the demand function $(1)($Equation $1)$

just given is the truth.

b$)$ Suppose widehat$({}_6)$ in Eq$.(1)$ turns out to be statistically insignificant. Does that mean there is no specification

error if we fit Eq$.(2)$ to the data?

c$)$ If widehat$({}_6)$ turns out to be insignificant, does that mean one should not introduce the price of a substitute

product$($s$)$ as an argument in the demand function?