Hi, I have some questions regarding Weighted Least Squares 1. Under what situation is it normally used? 2. Can I use 1/(Variance of Error) as the weight factor? 3. What are the things I need to check for model adequacy if I use this method? 4. Any recommended text for this topic? Lastly a question on prediction (related to economics): Say, I have obtained a model to predict expenditure from several towns /cities in a particular state (and I'm interest to actual predict the expenditure of Town X in the near future, with some variables held constant). As we know, based on the dataset to build the model, we have residual on the prediction for Town X's expenditure. How do I make use of this residual to adjust the prediction (since it's a mean value of several towns) of Town X's expenditure in the near future? If my explaination above is unclear, pls let me know and I'll try to elaborate more if possible. Thanks in advance! Regards, Beng Hai _________________________________________________________________________ Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com. Share information about yourself, create your own public profile at http://profiles.msn.com. ================================================================= Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =================================================================
