More about bell curves
Quote:
Originally Posted by wbroun
Well, one can see a very definite bell curve.

Only because of the varying interval size at the low end. If all intervals were $1000 increments my guess is that you would see a much more bimodal distribution with two peaks: a larger one at $1500 midinterval and a smaller one at $3500 midinterval. Since I don't have the raw data, I can't run it through my stats software to give you skew, kurtosis, or a Levine statistic for normalacy of curve. My guess is that the data suggests several things:
1. There is a threshold cost level for recreational skiing of about $450 (midpoint of second interval, only one response in first interval).
2. There is one group of skiers distributed around a mean expenditure of $1500 (midpoint of redistributed responses in $1000 intervalsclose to $1600 midpoint of current intervals).
3. There is a second group of skiers distributed around a mean expenditure of $3500 (midpoint of interval with second peak in distribution).
4. There is a statistically significant group of outlier responses above the $20000 level.
Or in other words:
1. Skiing is not free
2. Some skiers spend more than others (probably through more frequent equipment purchases, longer road trips, more expensive accommodation).
3. A small number of skiers spend oodles (scientific term there) of money and make resort owners very happy indeed.