Abstract
Once data are collected, edited, summarized and otherwise prepared for detailed analysis, basic methods of statistical inference are applied to address stated experimental and management objectives. In this chapter, we look at several key statistical techniques that are used in inference, particularly in the context of reliability and warranty analysis. These include (1) estimation, including maximum likelihood, several other methods of point estimation, and confidence intervals; (2) hypothesis testing, including comparison of two population means; (3) nonparametric methods for comparing populations; (4) tolerance intervals for estimating population fractiles; and (5) rank correlation for measuring data relationships.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
- 2.
In some cases, it is possible to use the data to check for evidence of bias. A better (but usually expensive) check is to randomly sample the entire population of sold items and compare key measures from that sample with those of the claims data.
- 3.
The chapter is intended to be a review of these topics, and, as such, treats most topics briefly, giving few details and no mathematical derivations. Details can be found in introductory statistical texts such as [13, 17]. Mathematical results and some techniques for their derivation are given in Appendix D; Details may be found in [6, 18].
- 4.
Simulation studies can often be used to investigate alternative methods in this situation, but with obvious limitations.
- 5.
The gamma function may be evaluated by interpolation in the tables found in [1], or by use of the gamma command in Minitab Calculator.
- 6.
- 7.
Note: In the Minitab parameterization of the exponential, the scale parameter is the inverse of that in the formula used in this book.
- 8.
- 9.
For a thorough discussion, see [19].
- 10.
- 11.
- 12.
- 13.
The estimators in this class are known as Best Asymptotically Normal (BAN) estimators. See [19] for details.
- 14.
- 15.
Note: most statistical program packages do not require the specification of a value for α as an input. Instead, the analysis is performed and the output includes a “p-value,” which is the probability of obtaining the observed value of the test statistic, given that H 0 is true. If p ≤ α, H 0 is rejected; if p > α, it is not.
- 16.
Note: The version of the z-statistic used for asymptotic tests may be replaced by the corresponding modification of the t-test for small sample sizes, say n < 100. In either case, the test is approximate in the sense that the level of significance is only approximately achieved. How good the approximation may be depends, among other things, on the true distribution of the test statistic and the sample size.
- 17.
- 18.
In general, an F-statistic is the ratio of two Chi-Square variables. df are those associated with the numerator and denominator of this ratio.
- 19.
- 20.
Note: Since Table E.3 gives values only to 100 df, the required fractiles of the Chi-Square distribution were obtained by use of Minitab.
- 21.
A key difficulty is that parametric tests applied to data from distributions very different from that assumed may result in error rates quite different from the nominal.
- 22.
Note: Minitab calculates the exact binomial probability of obtaining a result this or more discrepant from that expected under H 0, which may differ somewhat from that obtained using the normal approximation unless n is quite large.
- 23.
The test is sometimes described as a comparison of the medians of two populations, which is not strictly a correct interpretation.
- 24.
Note: The formula for the standard deviation assumes no ties. If there are only a few, this can safely be ignored. Otherwise, the formula requires an adjustment given in most texts on nonparametrics. Computer packages automatically perform the adjustment.
- 25.
See texts on nonparametric statistics for the tables. Additional details regarding the MannᾢWhitney U may be found in the Wikipedia article by that title and in [7].
References
Abramowitz M, Stegun IA (1964) Handbook of mathematical functions with formulas, graphs and mathematical tables. U.S. Government Printing Office, Washington
Blischke WR, Murthy DNP (2000) Reliability. Wiley, New York
Dixon WJ, Massey FJ Jr (1969) Introduction to Statistical Analysis, 3rd edn. McGraw-Hill, New York
Govindarajulu Z (1964) A supplement to Mendenhall’s bibliography on life testing and related topics. J Am Statist Assoc 59:1231ᾢ1291
Hirose H, Lai TL (1997) Inference from grouped data in three-parameter Weibull models with applications to breakdown-voltage experiments. Technometrics 39:199ᾢ210
Hogg RV, Craig A, McKean JW (2004) Introduction to mathematical statistics, 6th Edition edn. Prentice Hall, New York
Hollander M, Wolfe DA (1999) Nonparametric statistical methods, 2nd Edition edn. Wiley, New York
Høyland A, Raussand M (1994) System reliability theory. Wiley Interscience, New York
Johnson NL, Kotz S (1970) Distribution in statistics: continuous univariate distributionsᾢI. Wiley, New York
Kruskal W, Wallis A (1952) Use of ranks in one-criterion variance analysis. J Am Statist Assoc 47:583ᾢ621
Lawless JF (1982) Statistical models and methods for lifetime data. Wiley, New York
Martz HF, Waller RA (1982) Bayesian reliability analysis. Wiley, New York
Moore DS, McCabe GP, Craig B (2007) Introduction to the practice of statistics. W H Freeman, New York
Murthy DNP, Xie M, Jiang R (2004) Weibull models. Wiley, New York
Nelson W (1982) Applied life data analysis. Wiley, New York
Owen DB (1962) Handbook of statistical tables. Addison-Wesley, Reading, MA
Ryan TP (2007) Modern engineering statistics. John Wiley, New York
Somerville P (1958) Tables for obtaining non-parametric tolerance limits. Ann of Math Statist 29:599ᾢ601
Stuart A, Ord JK (1991) Kendall’s advanced theory of statistics, vol. 2, 5th Edition edn. Oxford University Press, New York
Wackerly D, Mendenhall W, Scheaffer RL (2007) Mathematical statistics with applications, Duxbury, New York 2007
Wilcoxon F (1945) Individual comparisons by ranking methods. Biometrics 1:80ᾢ83
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
Copyright information
© 2011 Springer-Verlag London Limited
About this chapter
Cite this chapter
Blischke, W.R., Rezaul Karim, M., Prabhakar Murthy, D.N. (2011). Basic Statistical Inference. In: Warranty Data Collection and Analysis. Springer Series in Reliability Engineering. Springer, London. https://doi.org/10.1007/978-0-85729-647-4_9
Download citation
DOI: https://doi.org/10.1007/978-0-85729-647-4_9
Published:
Publisher Name: Springer, London
Print ISBN: 978-0-85729-646-7
Online ISBN: 978-0-85729-647-4
eBook Packages: EngineeringEngineering (R0)