A Unified Framework for Bridging the Gap between Nonparametric and Parametric Inferences based on Ranking Data

Professor Yu, Philip Leung Ho   (Principal Investigator (PI))

Professor Alvo Mayer   (Co-Investigator)

42

2015-09-01

631972

A Unified Framework for Bridging the Gap between Nonparametric and Parametric Inferences based on Ranking Data

Block designs, Incomplete ranking, Nonparametric inference, Ranking data, Statistical modeling

Probability & Statistics

Physical Sciences (P)

17303515

General Research Fund (GRF)

2015

Completed

1 We will develop a new class of ranking models based on various choices of rank scores, and new nonparametric tests derived from the proposed models will be developed under different experimental designs. The model estimation procedures and power evaluation of the derived nonparametric tests will be studied. 2 We will extend the models to incorporate incomplete ranking data and covariates. When the dimension of the rank score vector is high, we may consider penalized likelihood estimation in order to narrow down the number of rank scores. 3 Applications to real-world ranking data sets will be considered during the development process of the proposed model and nonparametric methods. Software that will be useful for researchers will be developed.