Hi

Run the following spss program ... every time z is within 1.96 units of mu, mu 
is within lower and upper 95% CIs, as must be the case.  And both proportions / 
probabilities will be very close to the expected 95%, with slight variation 
across runs.

input program.
loop o = 1 to 100000.
end case.
end loop.
end file.
end input program.
compute mu = 100.
compute sigma = 15.
compute m = 0.
compute se = sigma/3.
compute z = 0.
compute lci = 0.
compute uci = 0.
compute pz = 0.
compute pci = 0.
do repeat v = v1 to v9.
compute v = rv.norm(mu,sigma).
end repeat.
compute m = mean(v1 to v9).
compute z = (m - mu)/se.
compute lci = m - 1.96*se.
compute uci = m + 1.96*se.
if abs(z) le 1.96 pz = 1.
if (mu ge lci) and (mu le uci) pci = 1.
freq pz.
freq pci.
crosstabs table pz by pci.

I'll leave it to better minds than mine to figure out why I cannot say that M 
and CI are both randomly selected from a hypothetical probability distribution, 
giving the (identical) probability assigned to each.

Take care
Jim


James M. Clark
Professor of Psychology
204-786-9757
204-774-4134 Fax
[email protected]


---
You are currently subscribed to tips as: [email protected].
To unsubscribe click here: 
http://fsulist.frostburg.edu/u?id=13090.68da6e6e5325aa33287ff385b70df5d5&n=T&l=tips&o=17431
or send a blank email to 
leave-17431-13090.68da6e6e5325aa33287ff385b70df...@fsulist.frostburg.edu

Reply via email to