Multilevel factorial designs with experiment-induced clustering

Inbal Nahum-Shani, John Dziak, Linda Marie Collins

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Factorial experimental designs have many applications in the behavioral sciences. In the context of intervention development, factorial experiments play a critical role in building and optimizing highquality, multicomponent behavioral interventions. One challenge in implementing factorial experiments in the behavioral sciences is that individuals are often clustered in social or administrative units and may be more similar to each other than to individuals in other clusters. This means that data are dependent within clusters. Power planning resources are available for factorial experiments in which the multilevel structure of the data is due to individuals' membership in groups that existed before experimentation. However, in many cases clusters are generated in the course of the study itself. Such experiment-induced clustering (EIC) requires different data analysis models and power planning resources from those available for multilevel experimental designs in which clusters exist prior to experimentation. Despite the common occurrence of both experimental designs with EIC and factorial designs, a bridge has yet to be built between EIC and factorial designs. Therefore, resources are limited or nonexistent for planning factorial experiments that involve EIC. This article seeks to bridge this gap by extending prior models for EIC, developed for single-factor experiments, to factorial experiments involving various types of EIC. We also offer power formulas to help investigators decide whether a particular experimental design involving EIC is feasible. We demonstrate that factorial experiments can be powerful and feasible even with EIC. We discuss design considerations and directions for future research.

Original languageEnglish (US)
Pages (from-to)458-479
Number of pages22
JournalPsychological Methods
Volume23
Issue number3
DOIs
StatePublished - Sep 1 2018

Fingerprint

Cluster Analysis
Research Design
Behavioral Sciences
Research Personnel
Power (Psychology)

All Science Journal Classification (ASJC) codes

  • Psychology (miscellaneous)

Cite this

@article{0c31abc1d36e401597d66f1f6a599665,
title = "Multilevel factorial designs with experiment-induced clustering",
abstract = "Factorial experimental designs have many applications in the behavioral sciences. In the context of intervention development, factorial experiments play a critical role in building and optimizing highquality, multicomponent behavioral interventions. One challenge in implementing factorial experiments in the behavioral sciences is that individuals are often clustered in social or administrative units and may be more similar to each other than to individuals in other clusters. This means that data are dependent within clusters. Power planning resources are available for factorial experiments in which the multilevel structure of the data is due to individuals' membership in groups that existed before experimentation. However, in many cases clusters are generated in the course of the study itself. Such experiment-induced clustering (EIC) requires different data analysis models and power planning resources from those available for multilevel experimental designs in which clusters exist prior to experimentation. Despite the common occurrence of both experimental designs with EIC and factorial designs, a bridge has yet to be built between EIC and factorial designs. Therefore, resources are limited or nonexistent for planning factorial experiments that involve EIC. This article seeks to bridge this gap by extending prior models for EIC, developed for single-factor experiments, to factorial experiments involving various types of EIC. We also offer power formulas to help investigators decide whether a particular experimental design involving EIC is feasible. We demonstrate that factorial experiments can be powerful and feasible even with EIC. We discuss design considerations and directions for future research.",
author = "Inbal Nahum-Shani and John Dziak and Collins, {Linda Marie}",
year = "2018",
month = "9",
day = "1",
doi = "10.1037/met0000128",
language = "English (US)",
volume = "23",
pages = "458--479",
journal = "Psychological Methods",
issn = "1082-989X",
publisher = "American Psychological Association Inc.",
number = "3",

}

Multilevel factorial designs with experiment-induced clustering. / Nahum-Shani, Inbal; Dziak, John; Collins, Linda Marie.

In: Psychological Methods, Vol. 23, No. 3, 01.09.2018, p. 458-479.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Multilevel factorial designs with experiment-induced clustering

AU - Nahum-Shani, Inbal

AU - Dziak, John

AU - Collins, Linda Marie

PY - 2018/9/1

Y1 - 2018/9/1

N2 - Factorial experimental designs have many applications in the behavioral sciences. In the context of intervention development, factorial experiments play a critical role in building and optimizing highquality, multicomponent behavioral interventions. One challenge in implementing factorial experiments in the behavioral sciences is that individuals are often clustered in social or administrative units and may be more similar to each other than to individuals in other clusters. This means that data are dependent within clusters. Power planning resources are available for factorial experiments in which the multilevel structure of the data is due to individuals' membership in groups that existed before experimentation. However, in many cases clusters are generated in the course of the study itself. Such experiment-induced clustering (EIC) requires different data analysis models and power planning resources from those available for multilevel experimental designs in which clusters exist prior to experimentation. Despite the common occurrence of both experimental designs with EIC and factorial designs, a bridge has yet to be built between EIC and factorial designs. Therefore, resources are limited or nonexistent for planning factorial experiments that involve EIC. This article seeks to bridge this gap by extending prior models for EIC, developed for single-factor experiments, to factorial experiments involving various types of EIC. We also offer power formulas to help investigators decide whether a particular experimental design involving EIC is feasible. We demonstrate that factorial experiments can be powerful and feasible even with EIC. We discuss design considerations and directions for future research.

AB - Factorial experimental designs have many applications in the behavioral sciences. In the context of intervention development, factorial experiments play a critical role in building and optimizing highquality, multicomponent behavioral interventions. One challenge in implementing factorial experiments in the behavioral sciences is that individuals are often clustered in social or administrative units and may be more similar to each other than to individuals in other clusters. This means that data are dependent within clusters. Power planning resources are available for factorial experiments in which the multilevel structure of the data is due to individuals' membership in groups that existed before experimentation. However, in many cases clusters are generated in the course of the study itself. Such experiment-induced clustering (EIC) requires different data analysis models and power planning resources from those available for multilevel experimental designs in which clusters exist prior to experimentation. Despite the common occurrence of both experimental designs with EIC and factorial designs, a bridge has yet to be built between EIC and factorial designs. Therefore, resources are limited or nonexistent for planning factorial experiments that involve EIC. This article seeks to bridge this gap by extending prior models for EIC, developed for single-factor experiments, to factorial experiments involving various types of EIC. We also offer power formulas to help investigators decide whether a particular experimental design involving EIC is feasible. We demonstrate that factorial experiments can be powerful and feasible even with EIC. We discuss design considerations and directions for future research.

UR - http://www.scopus.com/inward/record.url?scp=85017168540&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85017168540&partnerID=8YFLogxK

U2 - 10.1037/met0000128

DO - 10.1037/met0000128

M3 - Article

C2 - 28383950

AN - SCOPUS:85017168540

VL - 23

SP - 458

EP - 479

JO - Psychological Methods

JF - Psychological Methods

SN - 1082-989X

IS - 3

ER -