where t_((∝/2,v)) is the value of the Student’s t-distribution, with ν degrees of
freedom, that puts α/2 probability in the upper tail. [In the usual multiple
regression problem, ν = (n−p').] The standard error of (β_j ) ̂ is s ((β_j ) ̂) =√(c_jj s^2 )
where s^2 is estimated with ν degrees of freedom and c_jj is the (j + 1)th
diagonal element from 〖(X’X)〗^(-1).