The last poll on the topic "Are you a DJ?"....

[imported_Gman]

Found that djs outnumber non-djs two to one on this message board.

[ December 08, 2003, 10:35 AM: Message edited by: Gman ]

[Bill Blake]

The sample is biased.

You cant use that stat.

[JMJ]

OK, so based on the 3700 members currently registered, you're saying that 2464.2 of them are DJ's???.......JMJ graemlins/cool_shades.gif

['Magic' Juan]

Would bedroom DJ's count, or only those who play out in the big clubs? This, then, begs other questions.

magic_juan

[imported_Gman]

Originally posted by Jamie Lennox:
The sample is biased.

You cant use that stat. What do you mean ?

[imported_Gman]

Originally posted by JMJ:
OK, so based on the 3700 members currently registered, you're saying that 2464.2 of them are DJ's???.......JMJ graemlins/cool_shades.gif From the people who responded over time to the poll the percentages did not change after a while. If I extrapolate the percentages to the entire board membership then yes.


-G

[DJ RON C]

I knew we can take 'em in a fight. graemlins/cool_shades.gif

[Bill Blake]

Originally posted by Gman:
</font><blockquote>quote:</font><hr />Originally posted by Jamie Lennox:
The sample is biased.

You cant use that stat. What do you mean ? </font>[/QUOTE]Well, it has to be random (who becomes part of the sample population must be done so in a random way, and how you go about getting those people must be done is a random way).

For a population of the three thousand or so members, based on the total population, there is a formula to calculate how big the sample should be (which I forget) and then you could randomly send emails for that number and you would close to a ‘valid’ statistic
(as an example on how to get a good stat).

But sense the basis for the sample is just ‘those members that CHOSE to answer it’ (a topic post is hardly a way that gets you randomness) it doesn’t work because, for instance’ who could say that those members that chose to answer are a representation of the population of members?

Just some geeky shit.

[ December 08, 2003, 10:56 AM: Message edited by: Jamie Lennox ]

[Bill Blake]

Also, if you got EVERY member to answer the question, then you have something.

[imported_Gman]

Originally posted by Jamie Lennox:
</font><blockquote>quote:</font><hr />Originally posted by Gman:
</font><blockquote>quote:</font><hr />Originally posted by Jamie Lennox:
The sample is biased.

You cant use that stat. What do you mean ? </font>[/QUOTE]Well, it has to be random (who becomes part of the sample population must be done so in a random way, and how you go about getting those people must be done is a random way).

For a population of the three thousand or so members, based on the total population, there is a formula to calculate how big the sample should be (which I forget) and then you could randomly send emails for that number and you would close to a ‘valid’ statistic
(as an example on how to get a good stat).

But sense the basis for the sample is just ‘those members that CHOSE to answer it’ (a topic post is hardly a way that gets you randomness) it doesn’t work because, for instance’ who could say that those members that chose to answer are a representation of the population of members?

Just some geeky shit. </font>[/QUOTE]"...Just some geeky shit" is right.

graemlins/jpshakehead.gif Go back to work now and leave my stats alone.

[Bill Blake]

Originally posted by Gman:
</font><blockquote>quote:</font><hr />Originally posted by Jamie Lennox:
</font><blockquote>quote:</font><hr />Originally posted by Gman:
</font><blockquote>quote:</font><hr />Originally posted by Jamie Lennox:
The sample is biased.

You cant use that stat. What do you mean ? </font>[/QUOTE]Well, it has to be random (who becomes part of the sample population must be done so in a random way, and how you go about getting those people must be done is a random way).

For a population of the three thousand or so members, based on the total population, there is a formula to calculate how big the sample should be (which I forget) and then you could randomly send emails for that number and you would close to a ‘valid’ statistic
(as an example on how to get a good stat).

But sense the basis for the sample is just ‘those members that CHOSE to answer it’ (a topic post is hardly a way that gets you randomness) it doesn’t work because, for instance’ who could say that those members that chose to answer are a representation of the population of members?

Just some geeky shit. </font>[/QUOTE]"...Just some geeky shit" is right.

graemlins/jpshakehead.gif Go back to work now and leave my stats alone. </font>[/QUOTE]Hey basic stats is how I make a living...what do you expect?

[MusicFilter]

Originally posted by Gman:
Found that djs outnumber non-djs two to one on this message board. Of course everyone's a DJ these days. It's fashionable. mad1.gif