"Tim Smith" wrote in message
...
In article ,
kjpro @ usenet.com wrote:
But if brand X consistently outdoes brand Y in end-user results
surveys,
and it is not due to X actually being better than Y, then there is
*some* factor that is making the bad dealers favor Y. So, as a
consumer, I've got better odds going with brand X.


If you want to use inaccurate survey's to make your decisions, by all
means,
do so...

Can you point out how they are inaccurate? If a magazine surveys its
readers, and Brand X beats Brand Y by a statistically significant
margin, and it is not due to X actually being better, and it is not due
to poor installers favoring Y for some reason, then that would have to
mean that the readers of the magazine who pick X do a better job of
finding good installers, and the ones who pick Y are worse at picking
installers.

Why would that be? I suppose if X costs more than Y, then maybe
consumers who pick X are more careful when they pick their installers,
and get more references and recommendations, whereas the people buying
the cheaper brand might also tend to go with the cheapest installer?


Group 1
75 units of brand A was improperly installed.
25 calls were made to this group.

Group 2
75 units of brand B was improperly installed.
50 calls were made to this group.

Group 3
75 units of brand A was correctly installed.
50 calls were made to this group.

Group 4
75 units of brand B was correctly installed.
25 calls were made to this group.

Who wins the survey?
Is the survey accurate?
Does it help you in determining which brand is better?

What if brand A and B are exactly the same product?