Grubbs's test

Grubbs's test is based on the assumption of normality. That is, one should first verify that the data can be reasonably approximated by a normal distribution before applying the Grubbs test.^[2]

Grubbs's test detects one outlier at a time. This outlier is expunged from the dataset and the test is iterated until no outliers are detected. However, multiple iterations change the probabilities of detection, and the test should not be used for sample sizes of six or fewer since it frequently tags most of the points as outliers.^[3]

Grubbs's test is defined for the following hypotheses:

H₀: There are no outliers in the data set

H_a: There is exactly one outlier in the data set

The Grubbs test statistic is defined as

${\displaystyle G={\frac {\displaystyle \max _{i=1,\ldots ,N}\left\vert Y_{i}-{\bar {Y}}\right\vert }{s}}}$ 

with ${\displaystyle {\overline {Y}}}$  and ${\displaystyle s}$  denoting the sample mean and standard deviation, respectively. The Grubbs test statistic is the largest absolute deviation from the sample mean in units of the sample standard deviation.

This is the two-sided test, for which the hypothesis of no outliers is rejected at significance level α if

${\displaystyle G>{\frac {N-1}{\sqrt {N}}}{\sqrt {\frac {t_{\alpha /(2N),N-2}^{2}}{N-2+t_{\alpha /(2N),N-2}^{2}}}}}$ 

with t_{α/(2N),N−2} denoting the upper critical value of the t-distribution with N − 2 degrees of freedom and a significance level of α/(2N).

Grubbs's test can also be defined as a one-sided test, replacing α/(2N) with α/N. To test whether the minimum value is an outlier, the test statistic is

${\displaystyle G={\frac {{\bar {Y}}-Y_{\min }}{s}}}$ 

with Y_min denoting the minimum value. To test whether the maximum value is an outlier, the test statistic is

${\displaystyle G={\frac {Y_{\max }-{\bar {Y}}}{s}}}$ 

with Y_max denoting the maximum value.

Several graphical techniques can be used to detect outliers. A simple run sequence plot, a box plot, or a histogram should show any obviously outlying points. A normal probability plot may also be useful.

• Chauvenet's criterion
• Peirce's criterion
• Q test
• Studentized residual
• Tau distribution

• Grubbs, Frank (February 1969). "Procedures for Detecting Outlying Observations in Samples". Technometrics. 11 (1). Technometrics, Vol. 11, No. 1: 2–21. doi:10.2307/1266761. JSTOR 1266761.
• Stefansky, W. (1972). "Rejecting Outliers in Factorial Designs". Technometrics. 14 (2). Technometrics, Vol. 14, No. 2: 469–479. doi:10.2307/1267436. JSTOR 1267436.

  This article incorporates public domain material from the National Institute of Standards and Technology