Multilevel models and unbiased tests for group based interventions

Examples from the safer choices study

Scott C Carvajal, Elizabeth Baumler, Ronald B. Harrist, Guy S. Parcel

Research output: Contribution to journalArticle

19 Citations (Scopus)

Abstract

For many large-scale behavioral interventions, random assignment to intervention condition occurs at the group level. Data analytic models that ignore potential non-independence of observations provide inefficient parameter estimates and often produce biased test statistics. For studies in which individuals are randomized by groups to treatment condition, multilevel models (MLMs) provide a flexible approach to statistically evaluating program effects. This article presents an explanation of the need for MLM's for such nested designs and uses data from the Safer Choices study to illustrate the application of MLMs for both continuous and dichotomous outcomes. When designing studies, researchers who are considering group-randomized interventions should also consider the features of the multilevel analytic models they might employ.

Original languageEnglish (US)
Pages (from-to)185-205
Number of pages21
JournalMultivariate Behavioral Research
Volume36
Issue number2
StatePublished - 2001
Externally publishedYes

Fingerprint

Multilevel Models
Research Personnel
Nested Design
Group
Test Statistic
Biased
Assignment
statistics
Model
Estimate

ASJC Scopus subject areas

  • Mathematics (miscellaneous)
  • Statistics and Probability
  • Psychology(all)
  • Experimental and Cognitive Psychology
  • Social Sciences (miscellaneous)

Cite this

Multilevel models and unbiased tests for group based interventions : Examples from the safer choices study. / Carvajal, Scott C; Baumler, Elizabeth; Harrist, Ronald B.; Parcel, Guy S.

In: Multivariate Behavioral Research, Vol. 36, No. 2, 2001, p. 185-205.

Research output: Contribution to journalArticle

Carvajal, Scott C ; Baumler, Elizabeth ; Harrist, Ronald B. ; Parcel, Guy S. / Multilevel models and unbiased tests for group based interventions : Examples from the safer choices study. In: Multivariate Behavioral Research. 2001 ; Vol. 36, No. 2. pp. 185-205.
@article{0d4f773fb0674f4891641fbdd9b9a9f0,
title = "Multilevel models and unbiased tests for group based interventions: Examples from the safer choices study",
abstract = "For many large-scale behavioral interventions, random assignment to intervention condition occurs at the group level. Data analytic models that ignore potential non-independence of observations provide inefficient parameter estimates and often produce biased test statistics. For studies in which individuals are randomized by groups to treatment condition, multilevel models (MLMs) provide a flexible approach to statistically evaluating program effects. This article presents an explanation of the need for MLM's for such nested designs and uses data from the Safer Choices study to illustrate the application of MLMs for both continuous and dichotomous outcomes. When designing studies, researchers who are considering group-randomized interventions should also consider the features of the multilevel analytic models they might employ.",
author = "Carvajal, {Scott C} and Elizabeth Baumler and Harrist, {Ronald B.} and Parcel, {Guy S.}",
year = "2001",
language = "English (US)",
volume = "36",
pages = "185--205",
journal = "Multivariate Behavioral Research",
issn = "0027-3171",
publisher = "Psychology Press Ltd",
number = "2",

}

TY - JOUR

T1 - Multilevel models and unbiased tests for group based interventions

T2 - Examples from the safer choices study

AU - Carvajal, Scott C

AU - Baumler, Elizabeth

AU - Harrist, Ronald B.

AU - Parcel, Guy S.

PY - 2001

Y1 - 2001

N2 - For many large-scale behavioral interventions, random assignment to intervention condition occurs at the group level. Data analytic models that ignore potential non-independence of observations provide inefficient parameter estimates and often produce biased test statistics. For studies in which individuals are randomized by groups to treatment condition, multilevel models (MLMs) provide a flexible approach to statistically evaluating program effects. This article presents an explanation of the need for MLM's for such nested designs and uses data from the Safer Choices study to illustrate the application of MLMs for both continuous and dichotomous outcomes. When designing studies, researchers who are considering group-randomized interventions should also consider the features of the multilevel analytic models they might employ.

AB - For many large-scale behavioral interventions, random assignment to intervention condition occurs at the group level. Data analytic models that ignore potential non-independence of observations provide inefficient parameter estimates and often produce biased test statistics. For studies in which individuals are randomized by groups to treatment condition, multilevel models (MLMs) provide a flexible approach to statistically evaluating program effects. This article presents an explanation of the need for MLM's for such nested designs and uses data from the Safer Choices study to illustrate the application of MLMs for both continuous and dichotomous outcomes. When designing studies, researchers who are considering group-randomized interventions should also consider the features of the multilevel analytic models they might employ.

UR - http://www.scopus.com/inward/record.url?scp=0035535615&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0035535615&partnerID=8YFLogxK

M3 - Article

VL - 36

SP - 185

EP - 205

JO - Multivariate Behavioral Research

JF - Multivariate Behavioral Research

SN - 0027-3171

IS - 2

ER -