[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: You Be The Judge

Sorry for the possible duplication... the first instance of this reply
didn't show up.

>Somewhat controversially, at IBC Cintel ran a questionnaire, ...
>...A good statistician could, of course, argue with these figures, not least
>because the sample was of people who wanted to win a prize from Cintel!
>They are some indication, though, of the way people were thinking at IBC.
>David Fenton
>Cintel International

I am not a statistician (nor do I play one on TV) but I was taught some of
the basics.

What is the total number of respondents in the survey? On a small, sampled
survey the index of error is often 4 to 5%. This means that a 47/53% spread
may mean there is no real difference at all. I think it is remarkable that
so few of the questions uncovered a large enough difference to be

I thank Cintel for asking the questions and publishing the results. We are
all capable of biasing our own interpretations to fit individual needs.

--- David Tosh  dlt at earthlink.net
--- Video Engineer, Complete Post, Hollywood