Regression analysis in SPSS: identifying and managing heteroskedasticity (July 2019)
This video discusses how to assess your data for the presence of heteroskedasticity (i.e., non-constant variance in your regression results), as well as one strategy (use of heteroskedasticity-consistent estimator) for dealing with this problem in regression analysis. All demonstrations are carried out using SPSS. A copy of the Powerpoint can be downloaded here: drive.google.com/open?id=1Xyc... . A copy of the data can be downloaded here: drive.google.com/open?id=1hD7... .
For more videos on regression analysis, download a document containing a list of links here: drive.google.com/open?id=1tpq...
Thanks for watching!
Пікірлер: 8
This might not be the mostly watched vieo on youtube, but I am sure you saved everyone's life who watched it, mine for sure!
@mikecrowson2462
4 жыл бұрын
Hi there. Thanks for your comment! Glad you found it useful. Best wishes!
thank you!
Hi, Mike. I would like to ask why the studentized predicted value was used, and not the standardized predicted value. Thanks!
I am unable to views White test F test in my version 23
11:07 Hi, thanks for the great video. Just wondering how you come up with a threshold line below 4? It would be great if you could suggest me some references for value 4.
Binary coded variables should be entered in the fixed factors column ?
so if we're fixing the heteroscedasticity with robust standard errors and wanted to know partial/t-test f-test and the coefficient of determination, which result to use? the one with robust standards error or the regular linear regression? thanks in advance