| T O P I C R E V I E W |
| UKrieger |
Posted - 12/01/2021 : 07:55:07 AM
Origin Ver. 2017G b9.4.0.220
Operating System: Win 10
Dear colleagues,
I have a question concerning the behavior of the confidence interval when fitting data with prescribed instrumental error. If I set the instrumental error to be 10% of the data values, the fitting routine gives me a 95% confidence interval. But if I change this instrumental error (either to 1% or to 80%) this does not effect the confidence interval at all. However, if I prescribe no error, I get a much smaller confidence interval.
Do I make a mistake here?
Best, Uli Krieger |
| 1 L A T E S T R E P L I E S (Newest First) |
| YimingChen |
Posted - 12/01/2021 : 10:03:37 AM
The instrumental error comes into effect as weight for each data point in calculating confidence interval. So only the relative value matters. While if no error is set, that is equivalent to apply same weight to each point which will makes the result different.
James |
|
|
