Hi, I'm doing what is basic least square regression on 12 points (4 points in triplicate) but I get different results between excel and JMP. Can anyone offer an explanation? Thanks ResultExcelJMPIntercept-1.30885 -1.210003Slope0.9260180.9250579UCL (99.7%) at intercept3.11048410.0827...R20.99733670.997364 for data the data below: XY0 0.94200.15200.9182722.5972723.4642722.1845547.3295547.3295549.018137121.108137129.013137129.189

nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

Why do I get difference betwen excel and jmp when doing least square regression?

johnhart1

Hi,

I'm doing what is basic least square regression on 12 points (4 points in triplicate) but I get different results between excel and JMP. Can anyone offer an explanation? Thanks

Result	Excel	JMP
Intercept	-1.30885	-1.210003
Slope	0.926018	0.9250579
UCL (99.7%) at intercept	3.110484	10.0827...
R2	0.9973367	0.997364

for data the data below:

X	Y
0	0.942
0	0.152
0	0.918
27	22.597
27	23.464
27	22.184
55	47.329
55	47.329
55	49.018
137	121.108
137	129.013
137	129.189

Accepted answers

ron_horne

hi johnhart1,

I would guess you made a typo or missed and observation. I get the same results.

best,

ron

All comments

johnhart1

Quick addendum: I get the same result as excel when using the JMP excel pluggin but different when putting the numbers directly into JMP!

ron_horne

hi johnhart1,

I would guess you made a typo or missed and observation. I get the same results.

best,

ron

Kevin_Anderson

Friends don't let friends use Excel for statistics! There are many scholarly articles available about the statistical problems with Excel. Microsoft knows about them, yet won't fix them. Let the consumer beware.

mpb

In this case, my guess is that the data in Excel has greater precision then the 3 decimal places shown in the original post.

johnhart1

Hi,

I checked that bit and changed all the data to be strictly the same amount of decimal places

johnhart1

mmmhhhhh I get the same answer as well now....... something strange is going on somewhere.

Anyway! Thanks for the help!