Differences in causal estimates from longitudinal analyses of residualized versus simple gain scores: Contrasting controls for selection and regression artifacts

Robert E. Larzelere, Emilio Ferrer, Brett R Kuhn, Ketevan Danelia

Research output: Contribution to journalArticle

12 Citations (Scopus)

Abstract

This study estimates the causal effects of six corrective actions for children's problem behaviors, comparing four types of longitudinal analyses that correct for pre-existing differences in a cohort of 1,464 4- and 5-year-olds from Canadian National Longitudinal Survey of Children and Youth (NLSCY) data. Analyses of residualized gain scores found apparently detrimental effects of all corrective actions by parents and professionals on subsequent antisocial behavior and hyperactivity. In contrast, analyses of simple gain scores found only apparently beneficial effects. Temporally reversed analyses yielded the same pattern of results, consistent with selection biases and regression artifacts, not with unidirectional causal effects. The findings were similar for corrective actions by professionals (e.g., Ritalin, psychotherapy) and by parents (physical and nonphysical punishment, scolding/yelling, "hostile-ineffective" parenting). Longitudinal analyses should check for similar artifacts by implementing temporally-reversed analyses and by determining whether causally relevant coefficients would replicate without artifacts biased in their favor.

Original languageEnglish (US)
Pages (from-to)180-189
Number of pages10
JournalInternational Journal of Behavioral Development
Volume34
Issue number2
DOIs
StatePublished - 2010

Fingerprint

Artifacts
artifact
regression
Parents
Methylphenidate
parents
Punishment
Selection Bias
Parenting
Child Behavior
Psychotherapy
Longitudinal Studies
psychotherapy
penalty
trend
Problem Behavior

Keywords

  • Antisocial behavior
  • Causal inference
  • Growth curve analysis
  • Longitudinal analyses
  • Selection bias
  • Structural equation modelling

ASJC Scopus subject areas

  • Social Psychology
  • Education
  • Developmental and Educational Psychology
  • Social Sciences (miscellaneous)
  • Developmental Neuroscience
  • Life-span and Life-course Studies

Cite this

Differences in causal estimates from longitudinal analyses of residualized versus simple gain scores : Contrasting controls for selection and regression artifacts. / Larzelere, Robert E.; Ferrer, Emilio; Kuhn, Brett R; Danelia, Ketevan.

In: International Journal of Behavioral Development, Vol. 34, No. 2, 2010, p. 180-189.

Research output: Contribution to journalArticle

@article{d9ee73816e4143309794abb0f25b7cfa,
title = "Differences in causal estimates from longitudinal analyses of residualized versus simple gain scores: Contrasting controls for selection and regression artifacts",
abstract = "This study estimates the causal effects of six corrective actions for children's problem behaviors, comparing four types of longitudinal analyses that correct for pre-existing differences in a cohort of 1,464 4- and 5-year-olds from Canadian National Longitudinal Survey of Children and Youth (NLSCY) data. Analyses of residualized gain scores found apparently detrimental effects of all corrective actions by parents and professionals on subsequent antisocial behavior and hyperactivity. In contrast, analyses of simple gain scores found only apparently beneficial effects. Temporally reversed analyses yielded the same pattern of results, consistent with selection biases and regression artifacts, not with unidirectional causal effects. The findings were similar for corrective actions by professionals (e.g., Ritalin, psychotherapy) and by parents (physical and nonphysical punishment, scolding/yelling, {"}hostile-ineffective{"} parenting). Longitudinal analyses should check for similar artifacts by implementing temporally-reversed analyses and by determining whether causally relevant coefficients would replicate without artifacts biased in their favor.",
keywords = "Antisocial behavior, Causal inference, Growth curve analysis, Longitudinal analyses, Selection bias, Structural equation modelling",
author = "Larzelere, {Robert E.} and Emilio Ferrer and Kuhn, {Brett R} and Ketevan Danelia",
year = "2010",
doi = "10.1177/0165025409351386",
language = "English (US)",
volume = "34",
pages = "180--189",
journal = "International Journal of Behavioral Development",
issn = "0165-0254",
publisher = "SAGE Publications Ltd",
number = "2",

}

TY - JOUR

T1 - Differences in causal estimates from longitudinal analyses of residualized versus simple gain scores

T2 - Contrasting controls for selection and regression artifacts

AU - Larzelere, Robert E.

AU - Ferrer, Emilio

AU - Kuhn, Brett R

AU - Danelia, Ketevan

PY - 2010

Y1 - 2010

N2 - This study estimates the causal effects of six corrective actions for children's problem behaviors, comparing four types of longitudinal analyses that correct for pre-existing differences in a cohort of 1,464 4- and 5-year-olds from Canadian National Longitudinal Survey of Children and Youth (NLSCY) data. Analyses of residualized gain scores found apparently detrimental effects of all corrective actions by parents and professionals on subsequent antisocial behavior and hyperactivity. In contrast, analyses of simple gain scores found only apparently beneficial effects. Temporally reversed analyses yielded the same pattern of results, consistent with selection biases and regression artifacts, not with unidirectional causal effects. The findings were similar for corrective actions by professionals (e.g., Ritalin, psychotherapy) and by parents (physical and nonphysical punishment, scolding/yelling, "hostile-ineffective" parenting). Longitudinal analyses should check for similar artifacts by implementing temporally-reversed analyses and by determining whether causally relevant coefficients would replicate without artifacts biased in their favor.

AB - This study estimates the causal effects of six corrective actions for children's problem behaviors, comparing four types of longitudinal analyses that correct for pre-existing differences in a cohort of 1,464 4- and 5-year-olds from Canadian National Longitudinal Survey of Children and Youth (NLSCY) data. Analyses of residualized gain scores found apparently detrimental effects of all corrective actions by parents and professionals on subsequent antisocial behavior and hyperactivity. In contrast, analyses of simple gain scores found only apparently beneficial effects. Temporally reversed analyses yielded the same pattern of results, consistent with selection biases and regression artifacts, not with unidirectional causal effects. The findings were similar for corrective actions by professionals (e.g., Ritalin, psychotherapy) and by parents (physical and nonphysical punishment, scolding/yelling, "hostile-ineffective" parenting). Longitudinal analyses should check for similar artifacts by implementing temporally-reversed analyses and by determining whether causally relevant coefficients would replicate without artifacts biased in their favor.

KW - Antisocial behavior

KW - Causal inference

KW - Growth curve analysis

KW - Longitudinal analyses

KW - Selection bias

KW - Structural equation modelling

UR - http://www.scopus.com/inward/record.url?scp=77951046932&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77951046932&partnerID=8YFLogxK

U2 - 10.1177/0165025409351386

DO - 10.1177/0165025409351386

M3 - Article

AN - SCOPUS:77951046932

VL - 34

SP - 180

EP - 189

JO - International Journal of Behavioral Development

JF - International Journal of Behavioral Development

SN - 0165-0254

IS - 2

ER -