This is an old revision of this page, as edited by Mathstat (talk | contribs) at 16:28, 14 January 2011 (on recent tests). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.
Revision as of 16:28, 14 January 2011 by Mathstat (talk | contribs) (on recent tests)(diff) ← Previous revision | Latest revision (diff) | Newer revision → (diff)In statistics, normality tests are used to determine whether a data set is well-modeled by a normal distribution or not, or to compute how likely an underlying random variable is to be normally distributed.
More precisely, they are a form of model selection, and can be interpreted several ways, depending on one's interpretations of probability:
- In descriptive statistics terms, one measures a goodness of fit of a normal model to the data – if the fit is poor then the data is not well modeled in that respect by a normal distribution, without making a judgment on any underlying variable.
- In frequentist statistics statistical hypothesis testing, one tests the data against the null hypothesis that it is normally distributed.
- In Bayesian statistics, one does not "test normality" per se, but rather computes the likelihood that the data comes from a normal distribution with given parameters μ,σ (for all μ,σ), and compares that with the likelihood that the data comes from other distributions under consideration, most simply using Bayes factors (giving the relatively likelihood of seeing the data given different models), or more finely taking a prior distribution on possible models and parameters and computing a posterior distribution given the computed likelihoods.
Graphical methods
An informal approach to testing normality is to compare a histogram of the residuals to a normal probability curve. The actual distribution of the residuals (the histogram) should be bell-shaped and resemble the normal distribution. This might be difficult to see if the sample is small. In this case one might proceed by regressing the measured residuals against a normal distribution with the same mean and variance as the sample. If the regression produces an approximately straight line, then the residuals can safely be assumed to be normally distributed.
A more formal graphical tool is the normal probability plot, a quantile-quantile plot against the standard normal distribution. Here the correlation coefficient of the data (the goodness of fit of the best fit line) gives a measure of how well the data is modeled by a normal distribution. These also have the benefit that outliers stick out, and that they can be used for communication with non-statisticians more easily than numbers.
Back of the envelope test
A simple back-of-the-envelope test takes the sample maximum and minimum and computes their z-score, or more properly t-statistic (number of sample standard deviations that a sample is above or below the sample mean), and compares it to the 68–95–99.7 rule: if one has a 3σ event (properly, a 3s event) and significantly fewer than 300 samples, or a 4s event and significantly fewer than 15,000 samples, then a normal distribution significantly understates the maximum magnitude of deviations in the sample data.
This test is useful in cases where one faces kurtosis risk – where large deviations matter – and has the benefits that it is very easy to compute and to communicate: non-statisticians can easily grasp that "6σ events don’t happen in normal distributions".
Frequentist tests
Tests of univariate normality include D'Agostino's K-squared test, the Jarque–Bera test, the Anderson–Darling test, the Cramér–von-Mises criterion, the Lilliefors test for normality (itself an adaptation of the Kolmogorov–Smirnov test), the Shapiro–Wilk test, the Pearson's chi-square test, and the Shapiro–Francia test for normality.
Historically, the third and fourth standardized moments (skewness and kurtosis) were some of the earliest tests for normality. Mardia's multivariate skewness and kurtosis tests generalize the moment tests to the multivariate case. Other early test statistics include the ratio of the mean absolute deviation to the standard deviation and of the range to the standard deviation.
More recent tests of normality include the energy test (Szekely and Rizzo) and the tests based on the empirical characteristic function (ecf) (e.g. Epps and Pulley, Henze-Zirkler, BHEP tests). The energy and the ecf tests are powerful tests that apply for testing univariate or multivariate normality and are statistically consistent against general alternatives.
Bayesian tests
Kullback–Leibler distances between the whole posterior distributions of the slope and variance do not indicate non-normality. However, the ratio of expectations of these posteriors and the expectation of the ratios give similar results to the Shapiro–Wilk statistic except for very small samples, when non-informative priors are used.
Spiegelhalter suggests using Bayes factors to compare normality with a different class of distributional alternatives. This approach has been extended by Farrell and Rogers-Stewart.
Applications
One application of normality tests is to the residuals from a linear regression model. If they are not normally distributed, the residuals should not be used in Z tests or in any other tests derived from the normal distribution, such as t tests, F tests and chi-square tests. If the residuals are not normally distributed, then the dependent variable or at least one explanatory variable may have the wrong functional form, or important variables may be missing, etc. Correcting one or more of these systematic errors may produce residuals that are normally distributed.
Notes
- Judge et al. (1988) and Gujarati (2003) recommend the Jarque–Bera test.
- Mardia, K. V. (1970). Measures of multivariate skewness and kurtosis with applications. Biometrika 57, 519-530.
- (Filliben 1975)
- Szekely, G. J. and Rizzo, M. L (2005) A new test for multivariate normality, Journal of Multivariate Analysis 93, 58-80.
- Epps, T. W., and Pulley, L. B. (1983). A test for normality based on the empirical characteristic function. Biometrika 70, 723-726.
- Henze, N., and Zirkler, B. (1990). A class of invariant and consistent tests for multivariate normality. Communications in Statistics: Theory and Methods 19, 3595-3617.
- Henze, N., and Wagner, T. (1997). A new approach to the BHEP tests for multivariate normality. Journal of Multivariate Analysis 62, 1-23.
- Young K. D. S. (1993), Bayesian diagnostics for checking assumptions of normality. Journal of statistical computation and simulation, vol. 47, no 3–4, pp. 167–180
- Spiegelhalter, D.J. (1980). An omnibus test for normality for small samples. Biometrika, 67, 493–496. doi:10.1093/biomet/67.2.493
- Farrell, P.J., Rogers-Stewart, K. (2006) Comprehensive study of tests for normality and symmetry: extending the Spiegelhalter test. Journal of Statistical Computation and Simulation, 76(9), 803 – 816. doi:10.1080/10629360500109023
References
- Judge et al., Introduction to the Theory and Practice of Econometrics, Second Edition, 1988; 890–892.
- Filliben, J. J. (1975). "The Probability Plot Correlation Coefficient Test for Normality". Technometrics. 17 (1). American Society for Quality: 111–117. doi:10.2307/1268008.
{{cite journal}}: More than one of|number=and|issue=specified (help); Unknown parameter|month=ignored (help) - Gujarati, Damodar N., Basic Econometrics, Fourth Edition, 2003; 147–148
External links
- Implementation of the energy test of (multivariate) normality in R package energy].
- Information on the normality tests in the R package nortest.