Smartin,
Thank you so much for your insights!
As I am still learning about polynomial trending and its value, they are
helpful.
So, let me see if I understand what you are saying. Let's say, for instance,
we are talking about fluctuations in gas prices at the pump over the last x
number of years, and we see how they've gone up and down and up and down and
now are waaaaayyyy down (hallelujah! I paid $2.19 yesterday, a price I never
thought I'd see again in this lifetime), but we cannot predict future prices
based on trending of the data alone without knowing something about the
underlying context and 'reasons' for the fluctuations?
So, how is trending used then for predictive purposes? Does it merely state
probabilities of future performance based on previous fluctuations? If so,
what parameter reveals this probability (is it the R-value?)? So, using our
example above, could we plot a trend line that gives a probability, based on
the data, of gas prices going up or down, for instance?
Thank you, anyone, for enlightening me on this subject. Are there any
statisticians in our midst on this board?
Orion