[CQ-Contest] Packet use for single ops.

[DougKR2Q at aol.com]

In a message dated 11/13/00 11:31:08 AM Eastern Standard Time, 
lw9euj at ciudad.com.ar writes:

<< 78 % to allow packet for single ops.
 17 % to leave things the way they are.
 5 % not defined.
 
 Many of the guys who think that allowing packet for single ops is a bad
 thing, are ocasional contesters. While almost 90% of those who think that
 the use of packet by single ops would be OK, are well known hard-core
 contesters. >>

This is just not a valid statistical method.  You have not ensured a random 
sample, nor adequate coverage of population to be included.  While we in the 
USA are still waiting to see who our next #1 will be, we see the problems 
with "exit polling."  You need to determine what type contester is likely to 
respond to your question, whether or not you have used leading questions or 
"interviewer bias" and a myriad of other factors that absolutely will affect 
your "survey."

If you want to survey "hard core" contesters, you should do a survey of those 
who meet your definition.  You need to contact them.  YOU need to define the 
population...you cannot let the sample "define itself" by only counting 
"those who reply."  Non-response is a valid response.  Therefore, you may 
actually have only a 0.1% response...not a valid sample upon which to make 
inferences about the whole "hard core" contester population.  Also, by using 
numbers like "17%" or "78%" implies that you have had at least 100 
responses....I doubt that.  You level of precision is too great for the 
number of responses.

de Doug KR2Q


--
CQ-Contest on WWW:        http://lists.contesting.com/_cq-contest/
Administrative requests:  cq-contest-REQUEST at contesting.com