Quote:
Originally Posted by Phantopeth Does anyone know if when using the Data Analysis - Regression tool in Excel, it assumes homo- or heteroskadesticity? Thanks for any help. Random, I know. |
I imagine that Excel runs its regression homoscedastically by default, since you don't have to input individual point variances, though there's some uncertainty in that guess.
If you want to read why you shouldn't use Excel for running regressions or doing stats in the first place, read this
http://www.daheiser.info/excel/main/section9.pdf (really technical)
or this
http://www.cs.uiowa.edu/~jcryer/JSMTalk2001.pdf (not as much so, but still pretty technical)