0
Votes

What is bias?

By Lisa Kelly on the 11th of June, 2012

1 Answer

  • 0
    Vote

    A bias is a compensation value, a difference to an original value, which is added or subtracted. 

    Example: 
    The sine function returns values in the range of 1.0 to -1.0 
    If you don't want negative values but need to preserve the range between the highest and lowest value, here 2.0, you would add a bias of 1.0 to the result of the sine function.

    Refine By Pinaki Panda on the 16th of January, 2013

Suggested reading…

Find the upper and lower bounds of calculations

If we round 20.8cm to the nearest centimetre, it becomes 21cm.

If we are given the number 21cm and told it has been previously rounded to the nearest centimetre, we do not know what number it was originally.  For example it could have been 21.2cm, or 20.9cm, or 20.756cm.  What we do know, is the biggest (maximum) or smallest (minimum) number it could possibly have been.

The minimum value it could have been, is 20.5cm.  Any smaller, even a tiny bit, and it would have rounded down to 20cm, not up to 21cm.

The maximum value it could have been, is 21.499999999…cm.  Because this looks messy, we tend to just say the maximum value it could have been before rounding is 21.5cm.  Yes, this would have been rounded up to 22cm, but since the difference between 21.5cm and 21.499…cm is infinitely small anyway, and it looks so much neater, we say the maximum is 21.5cm.

If all the numbers in a calculation have been rounded, then there is a maximum and minimum value for the answer to the calculation.

e.g. 5m x 10m, where both measurements have been rounded to the nearest metre.

Maximum:  5.5m x 10.5m = 57.75m

Minimum: 4.5m x 9.5m = 42.75m

Nothing in this section yet. Why not help us get started?

Related Questions