Uncertainty and Best-Fit Lines
- Author:
- Lenore Horner
Each blue data point has 10 (tiny) simulated points with y-values chosen from the normal distribution
centered on the y-value of the blue dot and uncertainty given by the distance of the red dot from the blue dot. The line is a best fit line to all 3 blue points plus all 30 random points. Every change cause all the random points to be recreated. Cmd-R (on Mac) will also recreate the random points.
•Drag the red point at the top of any error bar error to change the standard deviation of that data point.
•Drag the blue circle to move a the center values of a data point.
•If you change the value of a data point with large standard deviation, how much does that affect the best fit line? Does the x-value of the point affect this?
•If you change the value of a data point with small standard deviation, how much does that affect the best fit line? Does the x-value of the point affect this?