325-328 ........ Discursive introduction. 329-340 ...... Cellular phone subscribers. [Cellular.jmp] 329 ........ Time series plot of raw data. 330 ........ The 1/4 power transform for subscribers. 331 ........ Residual plot shows the problem -- tracking (waves, cycles). 332 ........ Plotting the relationship between consecutive residuals. 333 ........ The Durbin-Watson statistic. 334 ........ A model with the lagged residuals. 335 ........ An improved (but imperfect) residual plot. 336 ........ Adding the lag 2 residuals -- a better looking residual plot. 337 ........ Looking at the model fits. 338 ........ Zooming in on the recent past -- a slight difference. 339 ........ Looking at prediction intervals. 340 ........ Summary. 341-358 ...... Trends in computer sales. [CompSale.jmp] 342-343 ..... Exploratory analysis. 344-345 ..... Basic models. 346-347 ..... Collinearity and autocorrelation problems. 348 ........ The Durbin-Watson statistic. 349 ........ The rationale for modeling differences. 350-353 ..... Model fitting for the differences. 354-355 ..... Checking the assumptions and leverage plots. 356 ........ A good looking residual plot. 357 ........ Summary. 358 ........ Quick and dirty prediction. Lagged response + trend.