Re: OT - googling for barometric pressure

Julia Altshuler wrote:
Melinda Shore wrote:

This one should be easy-peasy, nevertheless, if you're
reasonably disciplined about data collection. Here's an
online book on multivariate analysis:
http://www.psychstat.missouristate.edu/multibook/mlt00.htm

and an extremely excellent open-source stats package to do
the computation for you (and even draw pitchers):
http://www.r-project.org/


Easy-peasy? Perhaps you overestimate my poor variable awashed brain. Still, one has to start somewhere. I'll start data collection, then attack analysis. Thanks.

You can try the eyeball method. For each of the variables that you're interested in (heat, cold, rain, exercise on prior day, low pressure, etc.), make a little 2x2 matrix that has yes/no for the variable and yes/no for pain, and tally up the number of days where you had a mark in each cell, e.g.,

Pain
Low Pressure Yes No
Yes 8 1
No 2 7

Now, if you find a matrix where you find most of the days represented on a diagonal, like the one above, that would look like you may have a correlation, and you'd want to pursue it, gather more data.

If you get a matrix where you have nearly equal numbers in each of the cells, then you don't have any evidence of a correlation, e.g.,

Pain
Low Pressure Yes No
Yes 5 3
No 4 4

FurPaw

--
My family values don't involve depleted uranium.

To reply, unleash the dog.
.

Relevant Pages

  • Re: OT - googling for barometric pressure
    ... online book on multivariate analysis: ... and an extremely excellent open-source stats package to do ... I'll start data collection, then attack analysis. ...
    (rec.pets.dogs.behavior)
  • Re: Flaxseed oil greatly increases premature birth risk
    ... data collection. ... a very strong correlation: flaxseed oil. ... the average rate of premature births is 2 ... But for women consuming flaxseed oil in their last two ...
    (sci.med.nutrition)
  • Re: Insane Iislamic Bull Shit
    ... grossly overstating the scope of the report I cited and my intentions ... method of data collection. ... That's not correlation. ...
    (rec.gambling.poker)