Alternatives to statistical decision trees in regulatory (eco-)toxicological bioassays

Research output: Contribution to journalArticleResearchpeer review

Authors

  • Felix M. Kluxen
  • Ludwig A. Hothorn

Research Organisations

External Research Organisations

  • ADAMA Deutschland GmbH
View graph of relations

Details

Original languageEnglish
Pages (from-to)1135-1149
Number of pages15
JournalArchives of toxicology
Volume94
Issue number4
Early online date19 Mar 2020
Publication statusPublished - Apr 2020

Abstract

The goal of (eco-) toxicological testing is to experimentally establish a dose or concentration–response and to identify a threshold with a biologically relevant and probably non-random deviation from “normal”. Statistical tests aid this process. Most statistical tests have distributional assumptions that need to be satisfied for reliable performance. Therefore, most statistical analyses used in (eco-)toxicological bioassays use subsequent pre- or assumption-tests to identify the most appropriate main test, so-called statistical decision trees. There are however several deficiencies with the approach, based on study design, type of tests used and subsequent statistical testing in general. When multiple comparisons are used to identify a non-random change against negative control, we propose to use robust testing, which can be generically applied without the need of decision trees. Visualization techniques and reference ranges also offer advantages over the current pre-testing approaches. We aim to promulgate the concepts in the (eco-) toxicological community and initiate a discussion for regulatory acceptance.

Keywords

    Assumption tests, Hazard characterization, Hazard identification, Pre-tests, Regulatory toxicology, Robust statistics

ASJC Scopus subject areas

Cite this

Alternatives to statistical decision trees in regulatory (eco-)toxicological bioassays. / Kluxen, Felix M.; Hothorn, Ludwig A.
In: Archives of toxicology, Vol. 94, No. 4, 04.2020, p. 1135-1149.

Research output: Contribution to journalArticleResearchpeer review

Kluxen FM, Hothorn LA. Alternatives to statistical decision trees in regulatory (eco-)toxicological bioassays. Archives of toxicology. 2020 Apr;94(4):1135-1149. Epub 2020 Mar 19. doi: 10.1007/s00204-020-02690-w
Kluxen, Felix M. ; Hothorn, Ludwig A. / Alternatives to statistical decision trees in regulatory (eco-)toxicological bioassays. In: Archives of toxicology. 2020 ; Vol. 94, No. 4. pp. 1135-1149.
Download
@article{b7490a562cec437d8e57bd756f5fa385,
title = "Alternatives to statistical decision trees in regulatory (eco-)toxicological bioassays",
abstract = "The goal of (eco-) toxicological testing is to experimentally establish a dose or concentration–response and to identify a threshold with a biologically relevant and probably non-random deviation from “normal”. Statistical tests aid this process. Most statistical tests have distributional assumptions that need to be satisfied for reliable performance. Therefore, most statistical analyses used in (eco-)toxicological bioassays use subsequent pre- or assumption-tests to identify the most appropriate main test, so-called statistical decision trees. There are however several deficiencies with the approach, based on study design, type of tests used and subsequent statistical testing in general. When multiple comparisons are used to identify a non-random change against negative control, we propose to use robust testing, which can be generically applied without the need of decision trees. Visualization techniques and reference ranges also offer advantages over the current pre-testing approaches. We aim to promulgate the concepts in the (eco-) toxicological community and initiate a discussion for regulatory acceptance.",
keywords = "Assumption tests, Hazard characterization, Hazard identification, Pre-tests, Regulatory toxicology, Robust statistics",
author = "Kluxen, {Felix M.} and Hothorn, {Ludwig A.}",
year = "2020",
month = apr,
doi = "10.1007/s00204-020-02690-w",
language = "English",
volume = "94",
pages = "1135--1149",
journal = "Archives of toxicology",
issn = "0340-5761",
publisher = "Springer Verlag",
number = "4",

}

Download

TY - JOUR

T1 - Alternatives to statistical decision trees in regulatory (eco-)toxicological bioassays

AU - Kluxen, Felix M.

AU - Hothorn, Ludwig A.

PY - 2020/4

Y1 - 2020/4

N2 - The goal of (eco-) toxicological testing is to experimentally establish a dose or concentration–response and to identify a threshold with a biologically relevant and probably non-random deviation from “normal”. Statistical tests aid this process. Most statistical tests have distributional assumptions that need to be satisfied for reliable performance. Therefore, most statistical analyses used in (eco-)toxicological bioassays use subsequent pre- or assumption-tests to identify the most appropriate main test, so-called statistical decision trees. There are however several deficiencies with the approach, based on study design, type of tests used and subsequent statistical testing in general. When multiple comparisons are used to identify a non-random change against negative control, we propose to use robust testing, which can be generically applied without the need of decision trees. Visualization techniques and reference ranges also offer advantages over the current pre-testing approaches. We aim to promulgate the concepts in the (eco-) toxicological community and initiate a discussion for regulatory acceptance.

AB - The goal of (eco-) toxicological testing is to experimentally establish a dose or concentration–response and to identify a threshold with a biologically relevant and probably non-random deviation from “normal”. Statistical tests aid this process. Most statistical tests have distributional assumptions that need to be satisfied for reliable performance. Therefore, most statistical analyses used in (eco-)toxicological bioassays use subsequent pre- or assumption-tests to identify the most appropriate main test, so-called statistical decision trees. There are however several deficiencies with the approach, based on study design, type of tests used and subsequent statistical testing in general. When multiple comparisons are used to identify a non-random change against negative control, we propose to use robust testing, which can be generically applied without the need of decision trees. Visualization techniques and reference ranges also offer advantages over the current pre-testing approaches. We aim to promulgate the concepts in the (eco-) toxicological community and initiate a discussion for regulatory acceptance.

KW - Assumption tests

KW - Hazard characterization

KW - Hazard identification

KW - Pre-tests

KW - Regulatory toxicology

KW - Robust statistics

UR - http://www.scopus.com/inward/record.url?scp=85082875445&partnerID=8YFLogxK

U2 - 10.1007/s00204-020-02690-w

DO - 10.1007/s00204-020-02690-w

M3 - Article

C2 - 32193567

AN - SCOPUS:85082875445

VL - 94

SP - 1135

EP - 1149

JO - Archives of toxicology

JF - Archives of toxicology

SN - 0340-5761

IS - 4

ER -