Regression and 95% confidence

[WK]

Hi All - (Have already posted this in "Around the Water Cooler") - Its HPLC data
Been preparing an excel spreadsheet using formula from "Statistics and Chemometrics for Analytical Chemistry" Miller and Miller. This is for unweighted regression line. Using the data in the book they get R2 = 0.9988 and 95% confidence limit of 0.64 for xo=8.09. I've set that up nicely - getting the results all as in the book.
Now when I substitute my HPLC data and get R2=1.0000 and put in a mid point area value I get xo= "an expected value" but a 95% confidence limit 2 or 3 times the value for xo ??
Please - has anyone got any suggestions?
Data is (x/y):
3.5781/59216 , 5.2484/87087 , 7.5246/124577 , 9.8123/162195 , 11.4862/189891 , 14.4847/239959 , 16.4214/271852 .
So I get xo=9.80 for y of 162195 but 95% confidence limit 23.44?? (LOD 25.60)
Any pointers gratefully received !
Best Regards WK

_________________
-- Drunkship of Lanterns ---

[aceto_81]

Maybe your calculations in excel are wrong?
If I plug in your data in R, I get
xo=9.798747 (9.766186-9.831307 with 95% confidence)
If you want, you can send me your excel sheet, and I will take a look.

Ace

[WK]

Thanks for the reply aceto_81.
I will check again and get back to you.
Did you use excel for your calculation?
Regards
WK

[WK]

Hi aceto_81,
I have just done this at work using the extended version of excel at work and
replicated your findings - I only have basic home premium version at home
where I initially set this up - but I must have had it initially correct - to get the book values.
I have changed PCs recently at home and "converted" the excel files.
Anyway thanks - at least I have it at work PC.
WK

[aceto_81]

Thanks for the reply aceto_81.
I will check again and get back to you.
Did you use excel for your calculation?
Regards
WK

I used R-project (www.r-project.org) for the calculations.

Ace