Economic facts are typically established by means of statistical inference, which is concerned with how data supporting general claims about the world deduced from economic models should be calculated. Normally statistical inference proceeds by either estimation (point or interval) or testing. There is no general agreement on how statistical inference should be performed, but in some common situations there is good agreement about the numerical results. Of the three main schools – named after Fisher; Neyman, Pearson and Wald (NPW); and Bayes – the Bayesian alone is logically coherent.
KeywordsAnalysis of variance Bayes’ theorem Causality in economics de Finetti, B. Estimation Exchangeability Fisher, R. A. Frequentist probability Interval estimates Least-squares estimates Lehmann, E. Likelihood Lindley, D. V. Log-likelihood functions Maximum likelihood Minimax principle Neyman, J. Nuisance parameters Pearson, E.S. Point estimates Probability distribution Savage, L. J. Significance tests Statistical inference Subjective probability Testing Wald, A.
- De Finetti, B. 1974–5. Theory of probability: A critical introductory treatment. 2 vols. London: Wiley. Translation from the 1970 Italian original by A. Machi and A. Smith.Google Scholar
- Fisher, R.A. 1925. Statistical methods for research workers. Edinburgh: Oliver & Boyd.Google Scholar
- Fisher, R.A. 1935. The design of experiments. Edinburgh: Oliver & Boyd.Google Scholar
- Lehmann, E.L. 1959. Testing statistical hypotheses. New York: Wiley.Google Scholar
- Ramsey, F.P. 1931. The foundations of mathematics and other logical essays. London: Kegan, Paul, Trench, Trubner.Google Scholar
- Savage, L.J. 1954. The foundations of statistics. New York: Wiley.Google Scholar