Geography 823 Time Series Analysis Seminar Problem Set #6 Due date: Monday or Wednesday November 28 and 30, 2005 This assignment is intended to encourage you to use some of the statistical methods that we have described this quarter in the evaluation of a time series dataset of your choice, one that you may be using for your research work. These can include initial evaluations of basic data characteristics (box and whisker, histograms), evaluation of its normality. The analysis could include other evaluations of your data including chi-square evaluations of expected vs observed outcomes in the data, tests of statistical significance of the differences between sample means and/or variances. Analyses can include correlation among time series, regression, and other methods such as evaluation of data autocorrelation. Some methods soon to be discussed include principal component analysis, spectral analysis (which many will find interesting because it identifies potential cycles and periodicities in time series), and time series filtering and smoothing. If you have not begun developed research projects and would not like to find some data on your own, please see me for some potential time series to evaluate. Results from time series using techniques that we have covered should be prepared in the form of overheads, powerpoint slides (bring your laptops), and handouts for a 10-15 minute class presentation on either November 28 or 30. Please have a version of any of the above to turn in to the instructor. No expansive text is necessary.