Max Factor Infinity
how do you factor in absolute min’s and max’s for dataset to make better normal dist. approximations?
As it stands, the normal probability is computed from -infinity to a z-score. But what if we are dealing with human height? Cant be anything less than x=0. Possible z-scores cant be below a certain point. And yet we base our decisions on these obviously fallacious statistical reasoning
Maybe you are confusing Standard Normal Distribution with a Normal Distribution.
The Standard Normal Distribution has a mean of 0. So it has values to the left of 0.
If you are modeling human height as being normally distributed, the mean is not zero.
Go to this web site.
http://davidmlane.com/hyperstat/z_table.html
Say the mean height is 60 inches (5 feet) with a standard deviation of 3 inches.
You can find P( x < 55) It is 0.047790
You can also find P( x <0) It is 0.00000
You might also be noticing that Statistical Modeling does not give absolute answers. It is only as good at the model.
Someone might model human height might with a Normal Distribution. But maybe human height is not normally distributed. Then the answers from the model would be fallacious. The model needs to match the item being studied.
|
|
HEWLETT PACKARD PROCURVE 10gbe X2-CX4 Transceiver $2,082.48 The new design does not support the J8439A Optical Media Converter (OMC). The roll is needed only to identify which transceivers do and do not support the OMC…. |
|
|
Infinity Reference 1600a 600-Watt, High-Performance Mono Subwoofer Amplifier (Silver/Black) $374.95 The Reference 1600a features a 12dB per octave electronic crossover, allowing you to optimize bass performance based on enclosure type and subwoofer selection. Full-range preamp outputs faciliate use of multiple amplifiers without the need for noisy and expensive adapters or signal splitters. Port for future expansion products provided. Power Output: 400 watts RMS x 1 channel at 4 ohms and ? 1% T… |