New Tests for the Geometric Distribution
Author Information
Author(s): Coronel-Brizio Héctor Francisco, Hernández-Montoya Alejandro Raúl, Rodríguez-Achach Manuel Enrique, Tapia-McClung Horacio, Trinidad-Segovia Juan Evangelista
Primary Institution: Instituto de Investigaciones en Inteligencia Artificial, Universidad Veracruzana, Xalapa, Veracruz, México
Hypothesis
Can the Anderson-Darling and Watson tests be adapted for discrete models, specifically the geometric distribution?
Conclusion
The study successfully develops new goodness-of-fit tests for the geometric distribution, demonstrating their applicability in financial data analysis.
Supporting Evidence
- The tests developed provide a robust statistical framework for discrete distributions.
- Extensive tables of asymptotic critical values for the tests are included.
- The methodology is demonstrated through a financial case study involving major stock indices.
Takeaway
This study created new ways to check if data fits a specific pattern called the geometric distribution, which is useful for understanding trends in financial markets.
Methodology
The study adapts the Anderson-Darling and Watson tests for discrete distributions and applies them to financial time series data.
Limitations
The assumption of a constant probability of success over time may oversimplify the dynamic nature of financial markets.
Statistical Information
Confidence Interval
[0.490, 0.534]
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website