statistics percentage without a mean

This is not a homework or exam exercise, I would actually like to statistically compare some real data.

I would like to compare data collected from same source but by different people, and come up with some of percentage of variation in this method. Two people are measuring exactly the same thing, but coming up with different answers (human error). I would like to give this error a percentage of variance. (Is that what it is called?)

For an example, the two dip readings are 56 and 45, so the difference between them is 11. There is no mean or average. The angle of dip is at its maximum is 90 degrees, so how do I work out for a set of data what the average percentage error is between these two sets of measurements taken by different people.

In a same way, I would also like to compare the errors in the direction (strike) measurements. For example comparing a direction reading of 153 degrees (which is roughly SE) to 31 degrees (NE). Again, no mean value, only the knowledge that the is 360 degrees altogether.

I would be very happy to be even able to compare there measurements independently, but can you also compare the dip and direction variable in a same equation.

Phew....clearly one can say that I am no math person!

s