[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
RE: st: Stata vs SPSS
Surprisingly or not, I believe that
the attitude this person was expressing
is defensible, although I would naturally
express it differently.
It does nobody in statistical science any
good long term if the possibilities of prediction
from time series models are oversold.
For example, I guess that a standard marketing
question is predicting the likely sales of a new product.
In such cases, there is typically little (relevant) data,
a poor understanding of the underlying principles,
and very high sensitivity to future unknowns.
I doubt that the smartest analysts with the smartest
models can do much more than make a dent on an
intrinsically hard problem.
After all, Clive, what is your prediction for the results
of the next British election and how far is that
based on time series modelling?
> On this, I know whereof I speak. I remember having a casual webroom
> conversation with a lady who told me that she worked as a
> market analyst
> in the City of London (and probably still does). It went
> something like
> Me: "Oh really? So you try and predict market movements, do you?"
> Her: "Yup."
> Me: "Wow! So tell me; do you use time-series methods for prediction?"
> Her: "Er, nooooo."
> Me: "Um, OK. So what do you use, then?"
> Her: "Nothing other than our professional intuition. We stick
> a finger in
> our mouth, take it out, hold it up in the air, and
> judge which way the
> wind's blowing."
> Me: "Right. Don't you think you should be using something more, um,
> scientific, maybe?"
> Her: "No. Why should we when this works just as well? We've
> had lots of
> young, 'wet-behind-the-ears' kids come to us trying to
> teach us all
> this crap before, and we soon sent them packing."
> The discourse went downhill from there. With attitudes like this, no
> package that aims to be more comprehensive than MS Excel can
> hope to take
> a slice of the commerical cake, as it were, either here in the UK or
* For searches and help try: