# I Error in declination of linear regression

Tags:
1. May 21, 2017

### UiOStud

During a lab exercise we measured different masses of a magnetic material on a scale while changing the strength of the magnetic field it was in. Afterwards we plotted the masses and the fieldstrength hoping to find a linear slope. Then we drew a linear slope by using linear regression and found the declination of the slope and it's error. However this error does not take into account that the values of the masses also have an error. How can I fin the total error of the declination with the errorin the measurements in mind?

2. May 21, 2017

### FactChecker

Are you asking how you can remove the measurement error from your results to get the true slope of the line? I doubt that you can.

3. May 21, 2017

### UiOStud

Okay, let me try to explain in other words. I need to find the exact error in the linear slope I'm drawing. Because the points are not actually perfectly aligned there is an error in my regression. But finding the error from the residuals is not enough because there is an extra error comming from the fact that there is an error in the measurements themselves. The total error will be greater than the one I find only looking at the residuals. How can I find the total error of both the error in measurements and the error of regression?

4. May 21, 2017

### FactChecker

Regression algorithms in statistics packages will give you numbers for the statistical standard deviation of the estimated parameters, like the slope. They will not distinguish between different sources of variation. So the numbers they give will be for the total variation of the measurements and the mass.