J
Jack May
Guest
"John Kane" <[email protected]> wrote in message
news:[email protected]...
> On Jul 24, 4:39 pm, "Jack May" <[email protected]> wrote:
>> "Zoot Katz" <[email protected]> wrote in message
>>
>> news:[email protected]...
>>
>> > On Tue, 24 Jul 2007 16:01:11 +0000 (UTC), [email protected]
>>
>> > The model American puts in 1,600 hours to get 7,500 miles:
>> > less than five miles an hour."
>>
>> The Census says the average comute in th US is 12.1 miles and takes 22.5
>> minutes for an average speed of 32,3 MPH. Do you want to attempt
>> occasionally to tell truth?
>>
>> http://www-rcf.usc.edu/~pgordon/pdf/commuting.pdf
>> (page 4)
>
> Jack
> The problem here is that the authors apparently are using the
> arethmetic mean. Unless the distribution is normal the mean is almost
> certainly inflated by outliers. Do you know if anyone has done this
> type of study using median commuting distance rather than mean ?
Almost all real world statistics are non-Gaussian and the average is a poor
indicator. The table says the data uses averages
>
> in Canada the median commute is about 7.5 km . The Canadian
> situtation would seem quite different if you took the mean. If you
> look at the actual distribution in Canada the majority of commuters
> travel less than 10 km (6.2 miles). See
> http://ca.geocities.com/jrkrideau/cycling/commute.png.
> I would not be terribly surprised to find a similar distribution,
> althougth, perhaps with a slightly larger median value for US
> commutes.
Most real world statistics are power law which means the statistics for
different countries will be similar.
news:[email protected]...
> On Jul 24, 4:39 pm, "Jack May" <[email protected]> wrote:
>> "Zoot Katz" <[email protected]> wrote in message
>>
>> news:[email protected]...
>>
>> > On Tue, 24 Jul 2007 16:01:11 +0000 (UTC), [email protected]
>>
>> > The model American puts in 1,600 hours to get 7,500 miles:
>> > less than five miles an hour."
>>
>> The Census says the average comute in th US is 12.1 miles and takes 22.5
>> minutes for an average speed of 32,3 MPH. Do you want to attempt
>> occasionally to tell truth?
>>
>> http://www-rcf.usc.edu/~pgordon/pdf/commuting.pdf
>> (page 4)
>
> Jack
> The problem here is that the authors apparently are using the
> arethmetic mean. Unless the distribution is normal the mean is almost
> certainly inflated by outliers. Do you know if anyone has done this
> type of study using median commuting distance rather than mean ?
Almost all real world statistics are non-Gaussian and the average is a poor
indicator. The table says the data uses averages
>
> in Canada the median commute is about 7.5 km . The Canadian
> situtation would seem quite different if you took the mean. If you
> look at the actual distribution in Canada the majority of commuters
> travel less than 10 km (6.2 miles). See
> http://ca.geocities.com/jrkrideau/cycling/commute.png.
> I would not be terribly surprised to find a similar distribution,
> althougth, perhaps with a slightly larger median value for US
> commutes.
Most real world statistics are power law which means the statistics for
different countries will be similar.