How Do You Calculate Overall Error Percentage Between Two Data Sets?

[Jameseyboy]

Hi,

I have 3 values for data set 1 (1a, 1b, 1c), and 3 values for data set 2 (2a, 2b, 2c). (assume set 1 are the ones being compared to)

I know that to work out the error percentage for 1a and 2a -- it is the difference (of 1a and 2a) divided by 1a x 100.

I can do the same with 1b and 2b, and 1c, and 2c.

However, I need to work out the OVERALL error percentage for the data set 2, compared to 1.

How can this be done?

Thanks,

James

 

[mathman]

Overall error is too vague. My guess (since I am not sure what it means) would be to use the averages for both data sets and compare.

 

[wabbit]

There is no unique way to do that, and it depends what you mean by "overall error" and what you use it for
- you could just take the average, taking sign into account
- or the average of the absolute values
- or (perhaps the more common meaning), the square root of the sum of squared errors
- or many others;).