Related Experiment Videos
Relationships among multiple task asymmetries. I. A critical review
1Department of Philosophy, Psychology, and Cognitive Science, Rensselaer Polytechnic Institute, Troy, NY 21280, USA. bolesd@rpi.edu
Brain and Cognition
|July 2, 1998
Summary
This study reviews evidence on lateral difference measures, finding that local relationships are supported, but a global relationship is unlikely due to existing null and negative correlations. Factor analysis best reveals these local connections.
Area of Science:
- Neuroscience
- Cognitive Psychology
- Psychometrics
Background:
- Previous studies on lateral difference measures yielded conflicting interpretations regarding global versus local relationships.
- Existing literature is debated, with some supporting global and local relationships, while others support local relationships only.
Purpose of the Study:
- To reconcile the literature on lateral difference measures by reviewing evidence from principal components, correlational, and factor analytic approaches.
- To critically evaluate the evidence for global versus local relationships among lateral difference measures.
Main Methods:
- Review of existing literature utilizing principal components analysis.
- Review of existing literature utilizing correlational analysis.
- Review of existing literature utilizing factor analytic approaches.
Main Results:
- Evidence suggests that a global relationship among all asymmetries is unlikely, as null and negative correlations exist between measures.
- All reviewed approaches consistently support the reality of local relationships among lateral difference measures.
- Factor analytic approaches provide the most detailed insights into local relationships and hemispheric processes.
Conclusions:
- The existence of null and negative correlations challenges the concept of a singular global relationship in lateral asymmetry.
- Local relationships are consistently supported across multiple analytical methods, suggesting distinct underlying processes.
- Factor analytic methods offer the most robust framework for understanding local relationships, but require methodological improvements such as larger sample sizes, reliability prescreening, and data transparency.