Summary
We describe the way in which design and experimental error interact to determine convergence rates in central limit theorems for regression estimators. For example, we show that if the convergence rate in a central limit theorem for experimental errors alone isn −α, wheren is sample size and 0<α<1/2, then this rate is maintain in a central limit theorem for intercept and slope parameters if and only if the distribution generating design has finite (2+2α)'th moment. We prove that in other circumstances a careful choice of design can substantially improve convergence rates by introducing a degree of symmetry not present in the error distribution. Other results on the relationship between design and error are also derived.
References
Bhattacharya, R.N., Ranga Rao, R.: Normal approximation and asymptotic expansions. New York: Wiley 1976
Hall, P.: Rates of convergence in the central limit theorem. London: Pitman 1982
Hall, P., Barbour, A.D.: Reversing the Berry-Esseen inequality. Proc. Am. Math. Soc.30, 107–110 (1984)
Hall, P., Wightwick, J.C.H.: Two-sided bounds and leading terms for rates of convergence in the multivariate central limit theorem. J. Multivariate Anal.19, 225–239 (1986)
Heyde, C.C.: On the influence of moments on the rate of convergence to the normal distribution. Z. Wahrscheinlichkeitstheor. Verw. Geb.8, 12–18 (1967)
Heyde, C.C.: Some properties of metrics in a study of convergence to normality. Z. Wahrscheinlichkeitstheor. Verw. Geb.11, 181–192 (1969)
Heyde, C.C.: On the uniform metric in the context of convergence to normality. Z. Wahrscheinlichkeitstheor. Verw. Geb.25, 83–95 (1973)
Höglund, T.: On the convergence of convolutions of distributions with regularly varying tails. Z. Wahrscheinlichkeitstheor. Verw. Geb.15, 263–272 (1970)
Ibragimov, I.A.: On the accuracy of the approximation of distribution fof sums of independent random variables by the normal distribution. Theory Probab. Appl.11, 559–580 (1966)
Petrov, V.V.: Sums of independent random variables. Berlin Heidelberg New York: Springer 1975
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Hall, P. Effects of design and error on normal convergence rates in regression problems. Probab. Th. Rel. Fields 85, 283–305 (1990). https://doi.org/10.1007/BF01193941
Received:
Issue Date:
DOI: https://doi.org/10.1007/BF01193941
Keywords
- Stochastic Process
- Probability Theory
- Convergence Rate
- Experimental Error
- Limit Theorem