BreakTheChain.org

Crime Down Under

Tell me what you think about BreakTheChain.org. Comment on an article. Make a suggestion. Argue with me. Whatever. Don't hold back, I can take it.
Forum rules
If you are posting about a specific article on BreakTheChain.org, please provide the title and/or link of that article for reference.

Don't hold back, tell me what you think, but please be respectful of others' opinions and remember that everyone has the right to be wrong.

Crime Down Under

Postby AmosMoses on Wed Mar 04, 2009 2:36 am

You asked for it.

Regarding:

http://www.breakthechain.org/exclusives ... aguns.html

Your generalized statement that "the thing about statistics" is that "they can not show 'cause-and-effect'" is both irresponsible and somewhat asinine, considering the subject matter that your website addresses. To statistically show cause-and-effect accurately, one must simply run standardized and controlled testing where only the input variable is changed and the effect is measured.

I will grant you that historical data are generally too noisy for this sort of statistical analysis, but that's only because historical data is inherently an assay of data that is far too uncontrollable for accurate cause-and-effect analysis. Furthermore, depending on the assayed historical data, one can take these other factors into account and a quite accurate cause-and-effect analysis can be presented, the accuracy thereof dependent upon the knowledge and skill of the assayer -multiple skilled assayers being better, of course. Though possibly not "statistically sound", a very good and accurate analysis can be had with this sort of methodology. What's "statistically sound" is another whole subject in and of itself, one that is far beyond the scope of this posting to get into, and one that I probably couldn't accurately answer anyway, because I don't know jack shIt about statistics anyway...except to say that I have enough sense to know that sound cause-and-effect statistical analyses can definitely be made!

You CAN say accurately that "the thing about both of these sets of purported 'statistics'" is that "they most likely do not arrive at an accurate 'cause-and-effect' analysis" , however, your comments, as generally stated, deserve some chain breaking themselves.

You are welcome.
AmosMoses
 
Posts: 1
Joined: Wed Mar 04, 2009 2:05 am

Return to Site Feedback

Who is online

Users browsing this forum: Alexa [Bot], Yahoo [Bot] and 0 guests