"Rich Ulrich" <[EMAIL PROTECTED]> wrote in message [EMAIL PROTECTED]">news:[EMAIL PROTECTED]... > On Tue, 28 Jan 2003 19:14:05 GMT, "Mountain Bikn' Guy" <[EMAIL PROTECTED]> > wrote: > > [ snip, much... ]
> > X is autocorrelated, and Y is autocorrelated? Y is autocorrelated. The X variables are of many different data types, including nominal, ordinal, interval and continuous. Some of the continuous X's will show autocorrelation. Actually, our data set has thousands of independent variables, and I have not actually tested every one for autocorrelation. > [snip] > > Check for cross-correlation, and if there is not any, then > that changes the next questions. > I would not worry about applying diagnostics > until I spent some time in figuring whether the series > *ought* to be correlated, for some logical reason. > That should give me a hint of which critical > components need to be disentangled, which *changes* > ought to provide a sensitive test of co-variation. Theoretically, the autocorrelation of Y is appropriate but not meaningful in our models. Dave Reilly put is nicely: If the correct model is Y = 3X, then Y will be upward trending and will show autocorrelation. However, with the correct model, the causal variable X explains all the variability, and other observed values of Y are not needed in the model. In the absense of the correct X, autoregression would be helpful. > > -- > Rich Ulrich, [EMAIL PROTECTED] > http://www.pitt.edu/~wpilib/index.html . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
