Re: Re: Flawed Surveys [was: VPN endpoints]

["Bruce B. Platt" <bruce () ei3 com>]

Tina Bird wrote:


> I don't think that surveys are designed to be observations that test a
> theory.

You are rightly being quite precise here.

However, surveys can be used to test an hypothesis, which is  what I 
think Marcus and Paul are going back and forth on.

My favorite example of this is the MMPI (Minnesota Multiphasic 
Personality Inventory).  Several hundred T/F questions in length and 
administered to thousands of people, it's use was validated by looking 
at the clusters of score responses in the larger population and then 
comparing these to scores given by people who had independently been 
deemed to have some neurosis or psychosis by qualified practitioners.

I mention this for two reasons.  First is there a better question ever 
to ask someone than if they agree to the following:  "I am a secret 
agent of God."?     :-)

Second, if one allows this method as science, then there is a common 
ground for Paul and Marcus. As long as there is an agreed on validation 
method.  Isn't the lack of that the rub here?

Regards
_______________________________________________
firewall-wizards mailing list
firewall-wizards () honor icsalabs com
http://honor.icsalabs.com/mailman/listinfo/firewall-wizards