[Foundation-l] Is random article truly random

Andreas K. jayen466 at gmail.com
Fri Oct 21 05:06:29 UTC 2011


On Fri, Oct 21, 2011 at 2:48 AM, Bjoern Hoehrmann <derhoermi at gmx.net> wrote:

> * Andreas K. wrote:
> >The median and quartiles are on page 7 of the report:
> >
> >---o0o---
> >
> >Valid responses were received from  respondents between 10 – 85 years.
> >Overall, the average age of the Wikipedians that participated in the
> survey
> >is 25.22 years. Half of the respondents are younger than 22 years. The
> most
> >frequent age that can be observed within the respondents is 18
> >years. Splitting the respondents in four equally large age groups shows
> that
> >25% are younger than 18 years old, 25% are between 18 and 22, a further
> 25%
> >are between 22 and 30 (e.g. half of the respondents are between 18 and 30
> >years) and the remaining 25% are between 30 and 85 years old. There is a
> >slight age difference between readers and contributors - readers are, on
> >average, 24.79 years old while contributors show an average age of 26.14
> >years. Finally, female respondents are younger (23.79 years) than male
> ones
> >(25.69 years).
> >
> >---o0o---
>
> You made a point about editorial judgement and age, so I looked at data
> on editor age. As far as I can tell, the above only mentions the average
> age of "contributors", it does not say "The median age of contributors
> is 30 years" or some such thing.



The quoted text says that "half of the respondents are younger than 22
years". This is the same as saying that "the median age of respondents was
below 22 years of age", because that's how the median is defined.

The average age of contributors was higher than the average age of
respondents (readers + contributors) overall, but the difference was very
minor (1.35 years). So the median age of contributors wouldn't have differed
much either.

For the relative position of mode, median and average in a right-skewed
distribution see
http://en.wikipedia.org/wiki/File:Comparison_mean_median_mode.svg -- the
median is always smaller than the average.

Best,
Andreas



More information about the wikimedia-l mailing list