The Numerical Reliability of Econometric Software
Authors:
- 1 June 1999
- journal article
- review article
- Published by American Economic Association in Journal of Economic Literature
- Vol. 37 (2) , 633-665
Abstract
Numerous examples show that some econometric software packages contain serious flaws, and that users cannot safely assume that their software is accurate. A brief survey of the fundamentals of computer arithmetic discusses the sources of numerical error and emphasizes that computer arithmetic is not at all like pencil-and-paper arithmetic. Both users and developers of econometrics software should first pay attention to accuracy, and only later consider user-friendliness. Details are provided for assessing the accuracy of basic estimation routines, statistical distributions, and random number generators. More accuracy benchmarks are needed, especially for specialized econometric procedures.Keywords
This publication has 51 references indexed in Scilit:
- Certification of Algorithm 708ACM Transactions on Mathematical Software, 1994
- Measures of Fit for Calibrated ModelsJournal of Political Economy, 1993
- Remarks on my Term atJBESJournal of Business & Economic Statistics, 1993
- The market for (ir)reproducible econometricsSocial Epistemology, 1993
- Bias of some commonly-used time series estimatesBiometrika, 1983
- Test procedures and test problems for least squares algorithmsJournal of Econometrics, 1980
- The Acceptability of Regression Solutions: Another Look at Computational AccuracyJournal of the American Statistical Association, 1976
- Comparison of Several Algorithms for Computing Sample Means and VariancesJournal of the American Statistical Association, 1974
- An Appraisal of Least Squares Programs for the Electronic Computer from the Point of View of the UserJournal of the American Statistical Association, 1967
- Fourier Analysis of Uniform Random Number GeneratorsJournal of the ACM, 1967