Research Mentor(s)
Ramadha Piyadi Gamage
Description
Nonparametric tests are useful tools, as they can be used to analyze data where distributional assumptions cannot be made, and they often use a sorting or ranking system to do this. However, one thing to consider is that these methods often lead to ‘ties’ where data points cannot be ranked or sorted easily, and the method of dealing with such observations can affect the final results. The goal of this paper is to determine which method of tie-breaking was most useful for sign tests and Wilcoxon signed-rank tests. This usefulness was measured through the type I error rate and power of three different tie-breaking methods, when being used on generated data. For sign tests, this was done on uniform, binomial, and poisson-distributed data, and with sample sizes of 20, 50, and 100. Through these tests, this paper notes some significant differences between tie-breaking methods, as well as some trade-offs between them in more specific considerations, such as regarding distribution.
Document Type
Event
Start Date
May 2022
End Date
May 2022
Location
Carver Gym (Bellingham, Wash.)
Department
CSE - Mathematics
Genre/Form
student projects; posters
Type
Image
Rights
Copying of this document in whole or in part is allowable only for scholarly purposes. It is understood, however, that any copying or publication of this document for commercial purposes, or for financial gain, shall not be allowed without the author’s written permission.
Language
English
Format
application/pdf
Exploring Alternatives for Nonparametric Tests with Ties
Carver Gym (Bellingham, Wash.)
Nonparametric tests are useful tools, as they can be used to analyze data where distributional assumptions cannot be made, and they often use a sorting or ranking system to do this. However, one thing to consider is that these methods often lead to ‘ties’ where data points cannot be ranked or sorted easily, and the method of dealing with such observations can affect the final results. The goal of this paper is to determine which method of tie-breaking was most useful for sign tests and Wilcoxon signed-rank tests. This usefulness was measured through the type I error rate and power of three different tie-breaking methods, when being used on generated data. For sign tests, this was done on uniform, binomial, and poisson-distributed data, and with sample sizes of 20, 50, and 100. Through these tests, this paper notes some significant differences between tie-breaking methods, as well as some trade-offs between them in more specific considerations, such as regarding distribution.