- Sep 4, 2005
- 28,512
- 17,187
- Country
- United States
- Gender
- Male
- Faith
- Atheist
- Marital Status
- Single
- Politics
- US-Others
Try to remove Brazil, Mexico and South Africa from your dataset and plot homicide rate vs guns per 100k, and you'll find that r2 for that curve is 0.97 (that is it explains more of the variance). Is that correct? Probably not. But they are 3 outliers, and have alot of leverage. I could pick 3 other countries if you would like?
So basically, you want to toss 3 of the 12 countries that don't conform to the narrative, and replace them with 3 others that do?
Or is is something else you're asking?
Upvote
0