In a linear regression, a test of a simple linear hypothesis H₀ : f(θ‎₁,…,θ‎_K) = 0 against the alternative H₁ : f(θ‎₁,…,θ‎_K) ≠ 0 (two-tailed test) or H₁ : f(θ‎₁,…,θ‎_K) < 0 (one-tailed test), where θ‎ is a (K × 1) vector of regression parameters and f(·) is a scalar linear function. Under the null hypothesis, assuming normally distributed random errors, the test statistic, t=f(θ‎)/s.e.(f(θ‎)), has Student’s t-distribution with (N − K) degrees of freedom. Here θ‎ is the ordinary least squares estimator of θ‎ and s.e. is the standard error.