An adjusted bonferroni method for elimination of parameters in specification addition searches

Samuel B. Green, Marilyn Thompson, Jennifer Poirer

Research output: Contribution to journalArticle

16 Citations (Scopus)

Abstract

An adjusted Bonferroni method for controlling familywise error rate when deleting parameters in specification addition searches was presented and evaluated in a 2-part study. First, data were generated based on a factor model and then analyzed using backward selection, with and without applying the adjusted Bonferroni method. Three factors were manipulated: sample size, magnitude of weights, and number of parameters in a search. Under most of the 30 explored conditions, the empirical familywise error rates were relatively close to the nominal alpha level of. 05 when the adjusted Bonferroni method was applied. Error rates that exceeded the. 05 level appeared to be a function of the multivariate Wald test and not of the adjusted Bonferroni method. Second, data were generated based on a path model and analyzed using 2-stage and backward selection methods when the initial model was misspecified. Controlling stringently for Type I errors in the initial addition stage of a 2-stage search created selection errors when N was small (200). The adjusted Bonferroni method controlled adequately for Type I errors in the deletion stage of the 2-stage search and with backward selection.

Original languageEnglish (US)
Pages (from-to)18-39
Number of pages22
JournalStructural Equation Modeling
Volume8
Issue number1
DOIs
StatePublished - 2001

Fingerprint

Bonferroni
Elimination
Specification
Specifications
Familywise Error Rate
Type I error
Multivariate Tests
Wald Test
Factor Models
Deletion
Categorical or nominal
Error Rate
Sample Size
Path
Model

ASJC Scopus subject areas

  • Modeling and Simulation
  • Decision Sciences(all)
  • Economics, Econometrics and Finance(all)
  • Sociology and Political Science

Cite this

An adjusted bonferroni method for elimination of parameters in specification addition searches. / Green, Samuel B.; Thompson, Marilyn; Poirer, Jennifer.

In: Structural Equation Modeling, Vol. 8, No. 1, 2001, p. 18-39.

Research output: Contribution to journalArticle

@article{1a4ea846a9634554b95b7788e1e0b35d,
title = "An adjusted bonferroni method for elimination of parameters in specification addition searches",
abstract = "An adjusted Bonferroni method for controlling familywise error rate when deleting parameters in specification addition searches was presented and evaluated in a 2-part study. First, data were generated based on a factor model and then analyzed using backward selection, with and without applying the adjusted Bonferroni method. Three factors were manipulated: sample size, magnitude of weights, and number of parameters in a search. Under most of the 30 explored conditions, the empirical familywise error rates were relatively close to the nominal alpha level of. 05 when the adjusted Bonferroni method was applied. Error rates that exceeded the. 05 level appeared to be a function of the multivariate Wald test and not of the adjusted Bonferroni method. Second, data were generated based on a path model and analyzed using 2-stage and backward selection methods when the initial model was misspecified. Controlling stringently for Type I errors in the initial addition stage of a 2-stage search created selection errors when N was small (200). The adjusted Bonferroni method controlled adequately for Type I errors in the deletion stage of the 2-stage search and with backward selection.",
author = "Green, {Samuel B.} and Marilyn Thompson and Jennifer Poirer",
year = "2001",
doi = "10.1207/S15328007SEM0801_2",
language = "English (US)",
volume = "8",
pages = "18--39",
journal = "Structural Equation Modeling",
issn = "1070-5511",
publisher = "Psychology Press Ltd",
number = "1",

}

TY - JOUR

T1 - An adjusted bonferroni method for elimination of parameters in specification addition searches

AU - Green, Samuel B.

AU - Thompson, Marilyn

AU - Poirer, Jennifer

PY - 2001

Y1 - 2001

N2 - An adjusted Bonferroni method for controlling familywise error rate when deleting parameters in specification addition searches was presented and evaluated in a 2-part study. First, data were generated based on a factor model and then analyzed using backward selection, with and without applying the adjusted Bonferroni method. Three factors were manipulated: sample size, magnitude of weights, and number of parameters in a search. Under most of the 30 explored conditions, the empirical familywise error rates were relatively close to the nominal alpha level of. 05 when the adjusted Bonferroni method was applied. Error rates that exceeded the. 05 level appeared to be a function of the multivariate Wald test and not of the adjusted Bonferroni method. Second, data were generated based on a path model and analyzed using 2-stage and backward selection methods when the initial model was misspecified. Controlling stringently for Type I errors in the initial addition stage of a 2-stage search created selection errors when N was small (200). The adjusted Bonferroni method controlled adequately for Type I errors in the deletion stage of the 2-stage search and with backward selection.

AB - An adjusted Bonferroni method for controlling familywise error rate when deleting parameters in specification addition searches was presented and evaluated in a 2-part study. First, data were generated based on a factor model and then analyzed using backward selection, with and without applying the adjusted Bonferroni method. Three factors were manipulated: sample size, magnitude of weights, and number of parameters in a search. Under most of the 30 explored conditions, the empirical familywise error rates were relatively close to the nominal alpha level of. 05 when the adjusted Bonferroni method was applied. Error rates that exceeded the. 05 level appeared to be a function of the multivariate Wald test and not of the adjusted Bonferroni method. Second, data were generated based on a path model and analyzed using 2-stage and backward selection methods when the initial model was misspecified. Controlling stringently for Type I errors in the initial addition stage of a 2-stage search created selection errors when N was small (200). The adjusted Bonferroni method controlled adequately for Type I errors in the deletion stage of the 2-stage search and with backward selection.

UR - http://www.scopus.com/inward/record.url?scp=33846833527&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33846833527&partnerID=8YFLogxK

U2 - 10.1207/S15328007SEM0801_2

DO - 10.1207/S15328007SEM0801_2

M3 - Article

AN - SCOPUS:33846833527

VL - 8

SP - 18

EP - 39

JO - Structural Equation Modeling

JF - Structural Equation Modeling

SN - 1070-5511

IS - 1

ER -