How might one test circular data for symmetry? What asymmetric models are available for circular data should symmetry be rejected?

The talk will begin with a discussion of the problem of testing circular data for symmetry. In particular, the details of a new simple omnibus procedure based on large-sample theory will be considered. Pursuing the logical consequences of a rejection of symmetry, we will then move on to review briefly those asymmetric models which might be used in the analysis of circular data. The talk ends with an introduction to the wrapped skew-normal distribution on the circle and an illustration of its use in the finite mixture modelling of some data on bird migration.

Go back to Statistics seminars

Go to Department of Statistics home page