I spent too much time using my journalism-major R skills to mess with the Slate Star Codex spreadsheet, so I might as well tweet some stuff.
-
-
Replying to @RAVerBruggen
@RAVerBruggen A model I came up with picking variables and then doing the stepwise. Prob wrong but similar result.pic.twitter.com/VJDY2slCim
1 reply 0 retweets 0 likes -
Replying to @RAVerBruggen
@RAVerBruggen And here's when you pull out guns and then plot guns against the residuals.pic.twitter.com/xY5KPrDIb9
1 reply 0 retweets 0 likes -
Replying to @RAVerBruggen
@RAVerBruggen Q: To address multicollinearity (even mild) why isn't it standard to do this--remove variable of interest, use residuals?2 replies 0 retweets 0 likes -
Replying to @RAVerBruggen
@RAVerBruggen When I do that, coefficient falls quite a bit:pic.twitter.com/KqytN8maJ5
2 replies 0 retweets 0 likes -
Replying to @RAVerBruggen
@RAVerBruggen The coefficients are not very interpretable because they are not on the same scale. You need to standardize the data.2 replies 0 retweets 0 likes -
Replying to @KirkegaardEmil
@KirkegaardEmil You mean the residuals of a model are on a different scale than the original variable?2 replies 0 retweets 0 likes
@RAVerBruggen Yes, plot them and see.
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.