Submit Hint Search The Forums LinksStatsPollsHeadlinesRSS
14,000 hints and counting!


Click here to return to the 'Music and photos -- Summary graph of recent polls' hint
The following comments are owned by whoever posted them. This site is not responsible for what they say.
Music and photos -- Summary graph of recent polls
Authored by: roncross@cox.net on Aug 20, '04 12:52:12PM

Thanks for the graphs. Do you have the statistics that goes along with with?

For example, can you post the mean, median, standard deviation? All statistical graphs should have this kind of data with it.


If we exclude for a moment the data on the right after 3500-3999, where the bins are separated by a division of 1000, then we can see a skew distribution for both photos and songs prior to the bins labeled 4000-4999. Looking at the distribution from 0 to 3499 where the bins are separated by a division of 500 or less, the distribution is skewed as we would normally expect due to the boundary conditions set by the poll. For example, there is no way to have a negative number of photos and songs stored on your computer.

In the 4000 to 4999 bin, you see an increase in the response for songs and and photo, but this is somewhat misleading due to the fact that the bin has increased to 1000, which means that more stuff can be put in the bin. As you stay in the bins separated by 1000, the distribution starts to fall off again. If you want to compare the bin separated by a division of 500 with the bin separated by a division of 1000, then simply stack everything from none to 999, 1000 to 1999, 2000 to 2999, etc and you will get a good feel for the actual distribution. For example for songs, it would look like the following:

none to 999 would have roughly 626 responses
1000 to 1999 would have roughly 511
2000 to 2999 would have roughly 388
3000 to 3999 would have roughly 267
4000 to 4999 would have roughly 190

etc... you get to the point of diminishing return...

When you get to 10,000 or more, you are basically lumping the entire population after 10,000 and that is why it shoots up like a rocket!!! What this means is that there needs to be further division in polling. If you were to hypothetically separate out this bin to divisions of 500 or 1000, you will quickly find that it would just be a dribble of 0 to 1 response per bin. This is evident by the 171 and 270 responses in the songs and photo respectively where you would not have enough responses to but into a bin from 10000 to 11000 or even 10000 to 10500 for that matter.

thx
RLC

---
rlc



[ Reply to This | # ]
Music and photos -- Summary graph of recent polls
Authored by: wanny on Aug 21, '04 08:43:15PM

those are interesting histograms. The mean and standard deviation would be certainly be useful (perhaps skewness and kurtosis too -- which may help to answer rcl's question. Looks like the distribution is skewed to the left and is slightly leptkurtic to me (has long tails).

wanny



[ Reply to This | # ]
Music and photos -- Summary graph of recent polls
Authored by: gourls on Aug 23, '04 02:19:10PM

That was a great idea Rob. I did not take the poll, but now I wish I did because photos and songs are my favorite things on my iMac, besides Safari. Thanks.



[ Reply to This | # ]