Multilevel methods for modeling observed sequences of family interaction

George W. Howe; Getachew Dagne; C. Hendricks Brown

doi:10.1037/0893-3200.19.1.72

Multilevel methods for modeling observed sequences of family interaction

George W. Howe^*, Getachew Dagne, C. Hendricks Brown

^*Corresponding author for this work

Psychiatry and Behavioral Sciences

Research output: Contribution to journal › Article › peer-review

29 Scopus citations

Abstract

Observation of interaction plays a central role in family research. This article discusses how to analyze sequential data generated by discrete microcoding methods to test hypotheses about family interaction. Current methods for studying sequential data are presented, and their limits are discussed. Building on recent applications of contingency table analysis to such data, a multilevel log-linear model is presented that can specify and estimate indicators of individual behavioral tendencies and antecedent-consequent relationships among behaviors, both within and across samples of families. An example of this method is presented using data from a study of couples facing job loss. Potential extensions of this framework for future research are discussed.

Original language	English (US)
Pages (from-to)	72-85
Number of pages	14
Journal	Journal of Family Psychology
Volume	19
Issue number	1
DOIs	https://doi.org/10.1037/0893-3200.19.1.72
State	Published - Mar 2005

Keywords

Behavior observation
Empirical bayes
Random effects
Sequential analysis
Social interaction
Statistical methods

ASJC Scopus subject areas

General Psychology

Access to Document

10.1037/0893-3200.19.1.72

Cite this

@article{154409e51f644f47b18c29f95d569e29,

title = "Multilevel methods for modeling observed sequences of family interaction",

abstract = "Observation of interaction plays a central role in family research. This article discusses how to analyze sequential data generated by discrete microcoding methods to test hypotheses about family interaction. Current methods for studying sequential data are presented, and their limits are discussed. Building on recent applications of contingency table analysis to such data, a multilevel log-linear model is presented that can specify and estimate indicators of individual behavioral tendencies and antecedent-consequent relationships among behaviors, both within and across samples of families. An example of this method is presented using data from a study of couples facing job loss. Potential extensions of this framework for future research are discussed.",

keywords = "Behavior observation, Empirical bayes, Random effects, Sequential analysis, Social interaction, Statistical methods",

author = "Howe, {George W.} and Getachew Dagne and Brown, {C. Hendricks}",

year = "2005",

month = mar,

doi = "10.1037/0893-3200.19.1.72",

language = "English (US)",

volume = "19",

pages = "72--85",

journal = "Journal of Family Psychology",

issn = "0893-3200",

publisher = "American Psychological Association",

number = "1",

}

TY - JOUR

T1 - Multilevel methods for modeling observed sequences of family interaction

AU - Howe, George W.

AU - Dagne, Getachew

AU - Brown, C. Hendricks

PY - 2005/3

Y1 - 2005/3

N2 - Observation of interaction plays a central role in family research. This article discusses how to analyze sequential data generated by discrete microcoding methods to test hypotheses about family interaction. Current methods for studying sequential data are presented, and their limits are discussed. Building on recent applications of contingency table analysis to such data, a multilevel log-linear model is presented that can specify and estimate indicators of individual behavioral tendencies and antecedent-consequent relationships among behaviors, both within and across samples of families. An example of this method is presented using data from a study of couples facing job loss. Potential extensions of this framework for future research are discussed.

AB - Observation of interaction plays a central role in family research. This article discusses how to analyze sequential data generated by discrete microcoding methods to test hypotheses about family interaction. Current methods for studying sequential data are presented, and their limits are discussed. Building on recent applications of contingency table analysis to such data, a multilevel log-linear model is presented that can specify and estimate indicators of individual behavioral tendencies and antecedent-consequent relationships among behaviors, both within and across samples of families. An example of this method is presented using data from a study of couples facing job loss. Potential extensions of this framework for future research are discussed.

KW - Behavior observation

KW - Empirical bayes

KW - Random effects

KW - Sequential analysis

KW - Social interaction

KW - Statistical methods

UR - http://www.scopus.com/inward/record.url?scp=17244371889&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=17244371889&partnerID=8YFLogxK

U2 - 10.1037/0893-3200.19.1.72

DO - 10.1037/0893-3200.19.1.72

M3 - Article

C2 - 15796654

AN - SCOPUS:17244371889

SN - 0893-3200

VL - 19

SP - 72

EP - 85

JO - Journal of Family Psychology

JF - Journal of Family Psychology

IS - 1

ER -

Multilevel methods for modeling observed sequences of family interaction

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this