Recipe

Hypothesis testing procedure (t-test)

Use this to test a claim about one regression coefficient, for example "does income affect consumption?" or "is the marginal propensity to consume 80%?". It is the summary of Lecture 6's testing procedure.

  1. Write H0H_0 and H1H_1.
  2. Compute the t-statistic, t(β^1)=β^1−bse(β^1)t(\hat\beta_1)=\dfrac{\hat\beta_1-b}{se(\hat\beta_1)}.
  3. Get the critical value for the chosen α\alpha (1-tailed or 2-tailed row, correct df).
  4. Compare and decide. Equivalent shortcut: reject iff p<αp<\alpha (two-sided: iff b∉b\notin the (1−α)(1-\alpha) CI).

Common pitfalls

  • Say "fail to reject", never "accept": not rejecting only means the data are consistent with H0H_0.
  • The t value column of R's summary(lm) tests only H0:βk=0H_0:\beta_k=0; for any other null you must compute (β^k−b)/se(\hat\beta_k-b)/se yourself.

Worked example

Consumption on GDP (Israeli data, 32 df): β^1=0.81547\hat\beta_1=0.81547, se=0.01914se=0.01914, critical value t0.025[32]=2.0369t_{0.025}[32]=2.0369 at α=5%\alpha=5\%. To test whether the marginal propensity to consume is 80%, take H0:β1=0.8H_0:\beta_1=0.8 and compute t=(0.81547−0.8)/0.01914=0.808t=(0.81547-0.8)/0.01914=0.808. Since 0.808<2.03690.808<2.0369, fail to reject.