Can More Readings Reduce Fractional Error in Measurement?

[Angela Liang]

Why is fractional uncertainty not affected by systematic error? For example à vernier calipers measures the diameter of a coin:
(5.06+-0.04) mm
Can taking more readings, say 6, and taking average, reduce fractional error?

 

[FactChecker]

I'm not exactly sure about your terms, but here is my two cents:
A systemic error may be very consistent and predictable. That makes it conceivable to determine the error and make adjustments to the measured value and get the true value. On the other hand, a truly random error is difficult to determine and you can not make adjustments to the measured value.

A systemic error may just repeat the same error over and over, so taking the average of multiple readings will not reduce the error.

 

[Angela Liang]

I'm not exactly sure about your terms, but here is my two cents:
A systemic error may be very consistent and predictable. That makes it conceivable to determine the error and make adjustments to the measured value and get the true value. On the other hand, a truly random error is difficult to determine and you can not make adjustments to the measured value.

A systemic error may just repeat the same error over and over, so taking the average of multiple readings will not reduce the error.

Thanks!