Final Project - Stat 531 / Econ 677
The final project
should involve an analysis of the relationship between two or more time
series. The report is due on Tuesday, April 28 (the official final exam slot for this course) but may be
submitted earlier. The final project can be (but does not have to be) a
continuation of your midterm project. It must, however,
read as a
self-contained report; you may re-use any appropriate material.
The report should begin by presenting some detail concerning
your data and where you obtained it. Explain the scientific questions that
you are investigating. As for the midterm project, you should choose data
that interest you.
You should not knowingly use the same data as anyone else in
the class, though you should feel free to discuss your project.
The report should not assume that the reader is familiar with any specific software package.
In particular, you should not usually
present your computer code.
However you may comment on R commands, or other
computational issues, of particular interest.
It is expected that the report will contain
a graphical investigation of the data, consideration of second order
properties in both time (e.g. estimates of autocovariance, cross-covariance)
and frequency domain
(e.g. estimates of the spectrum,
cross-spectrum), a fitted model (which could
be, for example, a lagged regression model with ARMA errors),
some residual analysis and discussion of the results. It will probably be
necessary to look at the time series individually before studying their
relationship.
The report is to be no longer than 10 pages, including
figures. Points will be lost if it
exceeds that length. The report will be graded on the following two
considerations.
(i) Statistical methods: choice of methods, accuracy and clarity of the
statistical presentation, the capabilities and limitations of the
statistical methods.
(ii) Subject matter: motivation for the statistical
analysis, clearly stated goals and conclusions.
You have some flexibility whether to focus on the
subject matter addressed or on the statistical methods. Perhaps (i) may
be more appropriate for statisticians, and (ii) for economists,
scientists and engineers. Hopefully you'll be able to do a bit of both :)