By using this site, you agree to our Privacy Policy and our Terms of Use. Close
Kasz216 on 08 January 2013
Edit Thread
DaHuuuuuudge said:
Kasz216 said:
DaHuuuuuudge said:
Kasz216 said:
(Words, Oh so many words)


That's exactly what I said with more words.  Additionally I'm not argueing that more guns = less crime.   I'm argueing there is no correlation... at best their is a slight counter correlation to the opposite.

-_-

You peppered your sentences with condescension because Michael-5 was incorrect as to what statistical significance is, when you only cement my initial claim that you don't understand it either by making up terms such as "counter correlation".

sorry if I seem anal about this, I'm a statistician for the US gov't, and it really irks me when people get this shit wrong

cby ounter correlation i mean negative correlation.  Trying to put it into terms that he can understand.  Since he geniunely seems to be interested.   Like when people say "Reverse Racism" even though there is no such thing as "Reverse Racism" 

There generally tends to be a negative correlation.  Which is likely... completely unrelated.

As for there being no statistical significance... Look at the last link he posted.

What's the point of talking about Alpha when nobody knows what Alpha is?  (Note for others... see Correlation Coefficent) 

In general, you can often eyeball a linear regression trend line and know when it's not significant.  Unless someone built their graph in a stupid way stretching out either the X or Y.

-_-

The point of talking about alpha-levels is because essencially, they are what statistical significance is. You seem to be confusing correlation coefficient with alpha-levels, an alpha level tells you how likely it will be to reject the null hypothesis, and the correlation coefficient (r-value) shows the level of correlation as well as the nature of correlation (positive vs. negative)

You judge the correlation coefficent with the alpha....

I hate to go all wikipedia on you... which is weird that I'd have too since your a government stats guy but....

http://en.wikipedia.org/wiki/Statistical_significance

 

 

The significance level is usually denoted by the Greek symbol α (lowercase alpha). Popular levels of significance are 10% (0.1), 5% (0.05), 1% (0.01), 0.5% (0.005), and 0.1% (0.001). If a test of significance gives a p-value lower than the significance level α, the null hypothesis is rejected. Such results are informally referred to as 'statistically significant'. For example, if someone argues that "there's only one chance in a thousand this could have happened by coincidence", a 0.001 level of statistical significance is being implied. The lower the significance level chosen, the stronger the evidence required. The choice of significance level is somewhat arbitrary, but for many applications, a level of 5% is chosen by convention.[3][4]

 

 

If a correlation isn't strong enough, it can't pass the Alpha... because it's not strong enough to not be random chance.