Comment on “A New Statistic for Testing Suspected Outliers”

Douglas M. Hawkins

doi:10.1080/03610927708827503

Comment on “A New Statistic for Testing Suspected Outliers”

Douglas M. Hawkins

Statistics (Twin Cities)

Research output: Contribution to journal › Article › peer-review

10 Scopus citations

Abstract

A new statistic proposed by Tiku (1975) is compared with the Ferguson (1961) and Pearson-Chandra Sekar (1936) statistics. A simulation suggests that the new statistic is not superior to the Pearson-Chandra Sekar statistic when performance is assessed by the power of the test against the Dixon (1950) alternative In a recent paper, Tiku (1975) proposed a promising new family of statistics T for testing a normal sample for possible outliers. The family contains a test statistic corresponding to each r₁ and r₂, where r₁ is the number of outliers suspected on the left of the sample, and the number on the right. Other possible outlier statistics include the sample skewness b₁, (Ferguson 1961) Shapiro-Wilks W, and the Pearson-Chandra Sekar (1936) statistic L₁ = (X_(n) - X)/s. The first, two of these may be used to test for any number of outliers while LI is effective only for r₁, = 0 r₂ = 1.

Original language	English (US)
Pages (from-to)	435-438
Number of pages	4
Journal	Communications in Statistics - Theory and Methods
Volume	6
Issue number	5
DOIs	https://doi.org/10.1080/03610927708827503
State	Published - Jan 1 1977

Keywords

Dixon alternative
Pearson-Chandra Sekar statistic
Shaviro Wilks W
Tiku’s T statistic
extreme values

Access

10.1080/03610927708827503

OpenUrl availability

Full text

Cite this

@article{cbcd1cb32bad43dcb0572a6ee503bc3a,

title = "Comment on “A New Statistic for Testing Suspected Outliers”",

abstract = "A new statistic proposed by Tiku (1975) is compared with the Ferguson (1961) and Pearson-Chandra Sekar (1936) statistics. A simulation suggests that the new statistic is not superior to the Pearson-Chandra Sekar statistic when performance is assessed by the power of the test against the Dixon (1950) alternative In a recent paper, Tiku (1975) proposed a promising new family of statistics T for testing a normal sample for possible outliers. The family contains a test statistic corresponding to each r1 and r2, where r1 is the number of outliers suspected on the left of the sample, and the number on the right. Other possible outlier statistics include the sample skewness b1, (Ferguson 1961) Shapiro-Wilks W, and the Pearson-Chandra Sekar (1936) statistic L1 = (X(n) - X)/s. The first, two of these may be used to test for any number of outliers while LI is effective only for r1, = 0 r2 = 1.",

keywords = "Dixon alternative, Pearson-Chandra Sekar statistic, Shaviro Wilks W, Tiku{\textquoteright}s T statistic, extreme values",

author = "Hawkins, {Douglas M.}",

year = "1977",

month = jan,

day = "1",

doi = "10.1080/03610927708827503",

language = "English (US)",

volume = "6",

pages = "435--438",

journal = "Communications in Statistics - Theory and Methods",

issn = "0361-0926",

publisher = "Taylor and Francis Ltd.",

number = "5",

}

TY - JOUR

T1 - Comment on “A New Statistic for Testing Suspected Outliers”

AU - Hawkins, Douglas M.

PY - 1977/1/1

Y1 - 1977/1/1

N2 - A new statistic proposed by Tiku (1975) is compared with the Ferguson (1961) and Pearson-Chandra Sekar (1936) statistics. A simulation suggests that the new statistic is not superior to the Pearson-Chandra Sekar statistic when performance is assessed by the power of the test against the Dixon (1950) alternative In a recent paper, Tiku (1975) proposed a promising new family of statistics T for testing a normal sample for possible outliers. The family contains a test statistic corresponding to each r1 and r2, where r1 is the number of outliers suspected on the left of the sample, and the number on the right. Other possible outlier statistics include the sample skewness b1, (Ferguson 1961) Shapiro-Wilks W, and the Pearson-Chandra Sekar (1936) statistic L1 = (X(n) - X)/s. The first, two of these may be used to test for any number of outliers while LI is effective only for r1, = 0 r2 = 1.

AB - A new statistic proposed by Tiku (1975) is compared with the Ferguson (1961) and Pearson-Chandra Sekar (1936) statistics. A simulation suggests that the new statistic is not superior to the Pearson-Chandra Sekar statistic when performance is assessed by the power of the test against the Dixon (1950) alternative In a recent paper, Tiku (1975) proposed a promising new family of statistics T for testing a normal sample for possible outliers. The family contains a test statistic corresponding to each r1 and r2, where r1 is the number of outliers suspected on the left of the sample, and the number on the right. Other possible outlier statistics include the sample skewness b1, (Ferguson 1961) Shapiro-Wilks W, and the Pearson-Chandra Sekar (1936) statistic L1 = (X(n) - X)/s. The first, two of these may be used to test for any number of outliers while LI is effective only for r1, = 0 r2 = 1.

KW - Dixon alternative

KW - Pearson-Chandra Sekar statistic

KW - Shaviro Wilks W

KW - Tiku’s T statistic

KW - extreme values

UR - http://www.scopus.com/inward/record.url?scp=4544373527&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=4544373527&partnerID=8YFLogxK

U2 - 10.1080/03610927708827503

DO - 10.1080/03610927708827503

M3 - Article

AN - SCOPUS:4544373527

SN - 0361-0926

VL - 6

SP - 435

EP - 438

JO - Communications in Statistics - Theory and Methods

JF - Communications in Statistics - Theory and Methods

IS - 5

ER -

Comment on “A New Statistic for Testing Suspected Outliers”

Abstract

Keywords

Access

OpenUrl availability

Other files and links

Fingerprint

Cite this