Best Methods for Survival Data Analysis
Author Information
Author(s): Berkowitz Matthew, Altman Rachel MacKay, Loughin Thomas M.
Primary Institution: Statistics and Actuarial Science, Simon Fraser University, Burnaby, Canada
Hypothesis
Which methods for constructing survival trees and forests work best under various conditions?
Conclusion
The study identifies six top-performing methods for survival analysis and provides recommendations based on various factors affecting their performance.
Supporting Evidence
- Six methods were identified as top performers in predicting survival times.
- Factors like sample size and censoring significantly impact method performance.
- Recommendations were made for practitioners based on the study's findings.
Takeaway
This study looked at different ways to predict survival times and found some methods work better than others depending on the situation.
Methodology
An extensive simulation study was conducted to compare 11 methods for survival analysis based on various factors like sample size and censoring.
Potential Biases
Some methods may perform poorly under certain conditions, leading to biased predictions.
Limitations
The study's findings may not be generalizable to all survival datasets due to the specific conditions tested.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website