🚀 go-pugleaf

In article <5e89ms$817@rclnews.eng.ohio-state.edu> orlin-s@rclsgi.eng.ohio-state.edu (orlin steven jared) writes:




>In article <freematt-1602971720510001@cmh07.coil.com>,
>Matthew Gaylor <freematt@bronze.coil.com> wrote:
>>
>>Lake Research conducted a nationwide poll of 1,002 
>>Americans between February 5-9, 1997 for The Lindesmith 
>>Center, a Project of the Open Society Institute.  The 
>>survey was fielded by International Communications 
>>Research as part of the Excel Poll.  The following 
>>memorandum comes from Lake Research:

>A whopping 1002 Americans!!!!   I guess we can
>then say that 1002 Americans represents all of the 250,000,000
>people who live here!!  

We can we can make a number of assumptions but especially these two.
1.  The survey was of a properly randomized sample of Americans.  2.  The 
questions (which were listed verbatum in the post) actually relate to the 
claim.  The poster let us each judge the 2nd condition.  To say that 1002 
Americans cannot represent all 250,000,000 people who live here is a bit 
nieve.  It is true there is a margin of error that was not mentioned.

>>By a 2 to 1 margin, American voters favor allowing 
>>doctors to prescribe marijuana for medicinal purposes 
>>for seriously or terminally ill patients (60 percent 
>>favor, 30 percent oppose, nine percent don't know, and 
>>one percent refused to answer).

>So you've found 600 people who are for it, 300 people who
>are against it, and you call that "overwhelmingly" ???
>I call it a majority of a tainted survey.

Why do you think the survey was tainted?  It might have been.  And perhaps we 
should be skeptical of it since we (who know nothing of 
International Communications Research) know nothing of the 
methodolgy.  (And I am skeptical of it).  But to call it tainted?
 
Bert