Multilevel models for the experimental psychologist: Foundations and illustrative examples

Lesa Hoffman, Michael J. Rovine

Research output: Contribution to journalArticle

161 Citations (Scopus)

Abstract

Although common in the educational and developmental areas, multilevel models are not often utilized in the analysis of data from experimental designs. This article illustrates how multilevel models can be useful with two examples from experimental designs with repeated measurements not involving time. One example demonstrates how to properly examine independent variables for experimental stimuli or individuals that are categorical, continuous, or semicontinuous in the presence of missing data. The second example demonstrates how response times and error rates can be modeled simultaneously within a multivariate model in order to examine speed-accuracy trade-offs at the experimental-condition and individual levels, as well as to examine differences in the magnitude of effects across outcomes. SPSS and SAS syntax for the examples are available electronically.

Original languageEnglish (US)
Pages (from-to)101-117
Number of pages17
JournalBehavior Research Methods
Volume39
Issue number1
DOIs
StatePublished - Feb 2007

Fingerprint

Research Design
Theoretical Models
Psychology
Reaction Time
Experimental Psychologists
Experimental Design
Stimulus
Syntax
Response Time
Education
Categorical

ASJC Scopus subject areas

  • Experimental and Cognitive Psychology
  • Developmental and Educational Psychology
  • Arts and Humanities (miscellaneous)
  • Psychology (miscellaneous)
  • Psychology(all)

Cite this

Multilevel models for the experimental psychologist : Foundations and illustrative examples. / Hoffman, Lesa; Rovine, Michael J.

In: Behavior Research Methods, Vol. 39, No. 1, 02.2007, p. 101-117.

Research output: Contribution to journalArticle

Hoffman, Lesa ; Rovine, Michael J. / Multilevel models for the experimental psychologist : Foundations and illustrative examples. In: Behavior Research Methods. 2007 ; Vol. 39, No. 1. pp. 101-117.
@article{d6c109ac4c65400da9af3ea727802428,
title = "Multilevel models for the experimental psychologist: Foundations and illustrative examples",
abstract = "Although common in the educational and developmental areas, multilevel models are not often utilized in the analysis of data from experimental designs. This article illustrates how multilevel models can be useful with two examples from experimental designs with repeated measurements not involving time. One example demonstrates how to properly examine independent variables for experimental stimuli or individuals that are categorical, continuous, or semicontinuous in the presence of missing data. The second example demonstrates how response times and error rates can be modeled simultaneously within a multivariate model in order to examine speed-accuracy trade-offs at the experimental-condition and individual levels, as well as to examine differences in the magnitude of effects across outcomes. SPSS and SAS syntax for the examples are available electronically.",
author = "Lesa Hoffman and Rovine, {Michael J.}",
year = "2007",
month = "2",
doi = "10.3758/BF03192848",
language = "English (US)",
volume = "39",
pages = "101--117",
journal = "Behavior Research Methods",
issn = "1554-351X",
publisher = "Springer New York",
number = "1",

}

TY - JOUR

T1 - Multilevel models for the experimental psychologist

T2 - Foundations and illustrative examples

AU - Hoffman, Lesa

AU - Rovine, Michael J.

PY - 2007/2

Y1 - 2007/2

N2 - Although common in the educational and developmental areas, multilevel models are not often utilized in the analysis of data from experimental designs. This article illustrates how multilevel models can be useful with two examples from experimental designs with repeated measurements not involving time. One example demonstrates how to properly examine independent variables for experimental stimuli or individuals that are categorical, continuous, or semicontinuous in the presence of missing data. The second example demonstrates how response times and error rates can be modeled simultaneously within a multivariate model in order to examine speed-accuracy trade-offs at the experimental-condition and individual levels, as well as to examine differences in the magnitude of effects across outcomes. SPSS and SAS syntax for the examples are available electronically.

AB - Although common in the educational and developmental areas, multilevel models are not often utilized in the analysis of data from experimental designs. This article illustrates how multilevel models can be useful with two examples from experimental designs with repeated measurements not involving time. One example demonstrates how to properly examine independent variables for experimental stimuli or individuals that are categorical, continuous, or semicontinuous in the presence of missing data. The second example demonstrates how response times and error rates can be modeled simultaneously within a multivariate model in order to examine speed-accuracy trade-offs at the experimental-condition and individual levels, as well as to examine differences in the magnitude of effects across outcomes. SPSS and SAS syntax for the examples are available electronically.

UR - http://www.scopus.com/inward/record.url?scp=34347273771&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=34347273771&partnerID=8YFLogxK

U2 - 10.3758/BF03192848

DO - 10.3758/BF03192848

M3 - Article

C2 - 17552476

AN - SCOPUS:34347273771

VL - 39

SP - 101

EP - 117

JO - Behavior Research Methods

JF - Behavior Research Methods

SN - 1554-351X

IS - 1

ER -