Related Experiment Videos
SPSS and SAS programs for addressing interdependence and basic levels-of-analysis issues in psychological data.
1Department of Psychology, Lakehead University, 955 Oliver Road, Thunder Bay, ON, P7B 5E1 Canada. brian.oconnor@lakeheadu.ca
Behavior Research Methods, Instruments, & Computers : a Journal of the Psychonomic Society, Inc
|June 12, 2004
Summary
Levels-of-analysis issues occur when data from multiple individuals within the same group are analyzed, violating statistical assumptions. This study presents statistical programs to address these interdependence challenges in data analysis.
Area of Science:
- Social Sciences
- Psychology
- Organizational Behavior
Background:
- Individual-level data from interaction units (dyads, families, groups) present levels-of-analysis challenges.
- Interdependence among individuals within these units violates the independence-of-observations assumption in standard statistical tests.
Purpose of the Study:
- To describe data analysis challenges arising from levels-of-analysis issues.
- To present SPSS and SAS programs for appropriate statistical analyses.
- To contrast these methods with general multilevel modeling.
Main Methods:
- Description of data analysis challenges related to interdependence within interaction units.
- Implementation of within-and-between analyses (Dansereau et al., 1984).
- Implementation of dyad-level analyses (Gonzalez & Griffin, 1995, 1999).
Main Results:
- SPSS and SAS programs are provided to conduct appropriate statistical analyses for levels-of-analysis issues.
- The programs facilitate within-and-between analyses and dyad-level analyses.
- The study contrasts these specific methods with broader multilevel modeling approaches.
Conclusions:
- Appropriate statistical methods and software are crucial for analyzing data with interdependence.
- The presented programs offer practical solutions for researchers facing levels-of-analysis challenges.
- Understanding these specific analytical techniques enhances the validity of findings from grouped data.