Re: stat question

On Mar 22, 9:43 pm, st0...@xxxxxxxxxxxxx wrote:
There are tests of differences and tests of relationships. What are
examples of each and can you tell me when it might be used? I can't
seem to find information that makes sense to me.

It will never make sense to you because it does not make sense.
Standard statistical testing, and indeed much of standard statistics,
as taught in schoool and to unsuspecting pratictioners of various
sciences, is ad hoc and what is more, deeply flawed in many
circumstances to which it is supposed to apply. What is worse, this ad
hoc nature means that no one ever learns how to apply the techniques
learned to situations that they have not seen before; they must take
the 'advice' of an 'expert statistician'. These 'experts' never agree,
often not even on the basics, as you will see if you follow many of
the posts in this group.

Probability theory, however, (often called the Bayesian approach) does
not require all this ad hoc nonsense. It can be applied to all the
situations to which standard statistics is applied, is clear and
intuitive, assumptions are evident, and it is easy to generalize to
new situations. The effect of learning how to use it is summarized by
'give a man a fish and he can eat for a day, teach him how to fish and
he can eat for ever'.

For an easy and enjoyable introduction, read 'Probability theory: the
logic of science' by E. T. Jaynes. You can buy it (you will not regret
it) or you can download an old and incomplete version here:

http://omega.albany.edu:8008/JaynesBook.html

illywhacker;


.

Loading