Statistics: New t-tests for the comparison of two partially overlapping samples

Derrick, Ben

Abstract

This poster presented for the UWE degree show (June 2017), summarises Type I error rates and Power of standard t-tests and a new generalised form of the t-test, for comparing samples which include both paired and independent observations.

Materials are incorporated from a talk and presentation by the author at the Young Statisticians Meeting at the University College London (August 2016), assessing performance of the tests for Normal and non-normal distributions.

• Presentation Conference Type: Poster
• Conference Name: Faculty of Environment and Technology Degree Show
• Start Date: Jun 1, 2017
• End Date: Jun 1, 2017
• Acceptance Date: May 3, 2017
• Publicly Available Date: Jun 7, 2019
• Peer Reviewed: Not Peer Reviewed
• Keywords: type I error robustness, power, simulation, partially overlapping samples t-tests
• Public URL: https://uwe-repository.worktribe.com/output/888030
• Additional Information: Title of Conference or Conference Proceedings : Faculty of Environment and Technology Degree Show, UWE