5 ms·
> here's an even bigger problem to worry about: the government believes statisticians I think this is an oversimplification, and a dangerous one. These statist
by enko 15y ago
> here's an even bigger problem to worry about: the government believes statisticians
I think this is an oversimplification, and a dangerous one. These statistics are biased, one-sided and misleading. That is indeed a property of some or even most statistics, just as "being wrong" is a property of some or even most information. Yet no-one would think to proclaim that the problem with government is that it believes information.
Real statistics, like real science, is a force for good. Bad statistics, like pseudoscience, is a force for bad - but it can only fool the ignorant. The solution isn't to ban statistics. It's to educate decision makers so they can tell the difference.
- TeMPOraL 15y agoAgreed. I see lots of people arguing that statistics is a lie, a bad thing, a tool of evil, etc. Statistics is a tool - just like English language. I don't see people condemning English as a tool of evil because you can lie in it.
- eru 15y agoAbout English: http://en.wikipedia.org/wiki/E-Prime http://en.wikipedia.org/wiki/E-Prime
- RyanMcGreal 15y ago> It's to educate decision makers so they can tell the difference. The problem is not so much a lack of education as it is an excess of corporate influence.
- r00fus 15y agoAgreed. Any social program in the government uses stastical methods to determine utility, impact and ROI. Why? It works. Would you prefer a more "faith-based" approach? Perhaps in our day and age of more direct interactions (ie, email, SMS, twitter) it might be possible to augment statistical methods with direct feedback.