Stability of Multivariate Data Modeling in Longitudinal Data
Author Information
Author(s): Sengul Haydar, Barmada M Michael
Primary Institution: University of Pittsburgh
Hypothesis
How stable are multivariate models when applied to longitudinal data?
Conclusion
The study found good stability in predicted factor models and linkage analysis results over time.
Supporting Evidence
- Factor models showed high correlations across different time points.
- Linkage analysis results were consistent with previous studies.
- Factor analysis helped identify genetic factors related to cardiovascular disease.
Takeaway
The researchers looked at data over many years to see if their models stayed the same, and they found that they did a good job of staying stable.
Methodology
Factor analysis was applied to longitudinal data from the Framingham Heart Study, with adjustments for age and gender.
Limitations
Identifiability issues in factor models and potential numerical instability in early time points.
Participant Demographics
Participants were from the Framingham Heart Study, comprising 4692 subjects from large pedigrees.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website