Related Experiment Videos
Multilevel methods for modeling observed sequences of family interaction
George W Howe1, Getachew Dagne, C Hendricks Brown
1Department of Psychiatry and Behavioral Sciences, George Washington University, Washington, DC 20037, USA. cfrgwh@gwumc.edu
Summary
Analyzing family interaction data is crucial. This study presents a new multilevel log-linear model for examining behavioral sequences and individual tendencies in family research, offering advanced insights into interaction patterns.
Area of Science:
- Family Studies
- Quantitative Psychology
- Sociology
Background:
- Observing interaction is fundamental to family research.
- Analyzing sequential data from discrete microcoding methods is essential for hypothesis testing.
- Existing methods for sequential data analysis have limitations.
Purpose of the Study:
- To present a novel multilevel log-linear model for analyzing family interaction data.
- To address limitations in current sequential data analysis methods.
- To estimate individual behavioral tendencies and antecedent-consequent relationships.
Main Methods:
- Utilized contingency table analysis as a foundation.
- Developed and applied a multilevel log-linear model.
- Analyzed sequential data from discrete microcoding of family interactions.
Main Results:
- The proposed model can specify and estimate individual behavioral tendencies.
- The model effectively analyzes antecedent-consequent relationships within and across family samples.
- Demonstrated the method's utility with data from couples experiencing job loss.
Conclusions:
- The multilevel log-linear model offers a robust framework for analyzing complex family interaction patterns.
- This approach enhances the understanding of behavioral dynamics in families.
- The framework provides a basis for future research extensions in family interaction analysis.