I have sample analysis data from two seperate reports - the first done
to evaluate biased locations on a property and the second to validate
the findings of the first. As expected there are differences in the
values obtained, however, i lack the statiscal training to explain to
layman why those differences exist (standard deviation) and more
importantly i lack the ability to analyze them in a sound manner. I
need someone to receive the data and perform the statisical analysis
to demonstrate why the spread is to have been expected. The
non-science/math types think the numbers should have been exactly the
same. HELP

---

Clarification of Question by impact5438822-ga on 26 Feb 2006 11:23 PST

I should state that these are parts per billion results on soil samples

---

Request for Question Clarification by raisingmyhand-ga on 26 Feb 2006 12:39 PST

If you can provide margin of error information for the tests or
detailed information on the specific tests used and the results
obtained, researchers might be able to provide guidance on explaining
the differences and determining whether they are significant.

Try contact a company called pisces conservation in the UK. They come
up first in google. They specialise in statistical calcs and in
particular biological ones.

First of all you have two problems one a measuring problem especially
on ppm level. and the second one with normal distribution of data.
There are about (according to the bleack belt statistics handbook)100
different methods to calculate with the data. What I want from you is
two things 1) validation of measuring equipment for example if you
used AAS.I need blanc sample and samples for obtaining your data.
These are called reference measurements (For example 1 ,2, 3,4 5,etc
ppm for your measurements)Now the next step (step 2) is validate your
measurements serie 1 and serie 2 separate and 1 and 2 together. I wont
bother you wich method but if you mail me the data I will calculate
for you.
So here you are: HAVE NO FEAR I AM HERE
Prof kees

First of all you have two problems one a measuring problem especially
on ppb level. and the second one with normal distribution of data.
There are about (according to the bleack belt statistics handbook)100
different methods to calculate with the data. What I want from you is
two things 1) validation of measuring equipment for example if you
used AAS.I need blanc sample and samples for obtaining your data.
These are called reference measurements (For example 1 ,2, 3,4 5,etc
ppb for your measurements)Now the next step (step 2) is validate your
measurements serie 1 and serie 2 separate and 1 and 2 together. I wont
bother you wich method but if you mail me the data I will calculate
for you.
So here you are: HAVE NO FEAR I AM HERE
Prof kees