The pitfalls and profits of time series analysis

Abstract

The text considers the problems of isolating the various components of a time series and the effects these components have on each other. Also the idea of a periodogram is considered with a view to indicating the period for cyclical effect. Forecasting, using significance tests as a safeguard and contrasting moving average and least-squares trend lines, is reviewed in some detail. Data are used from the Wilde Theatre, South Hill Park, Bracknell, and also Pleasurama monthly share prices, to illustrate the ideas under discussion.