0

I 'm involved in a study where precentage values of carbon were take beneath between two adjacent fields. The adjacent fields differ by the vegetation cover (X, Y). Vegetation X is novel (planted 10 yrs ago) whereas vegetation Y has been there for at least 50 yrs.

I take the difference of carbon values by subtracting sample taken from soil beneath veg X and soil beneath veg Y.

Now, would it be more informative or meaningful if I take the relative or the absolute difference between the two samples? :

Relative difference: X-Y/Y. I would reason that this would make sense if I take into account that veg Y has been there "forever" and so the effect of veg X on carbon can be seen as a deviation from the potential effect of permanent vegetation Y. Then I coudl just argue like this: carbon under veg X has declined 1/3 with respect to veg Y.

What would be the information in this?

Absolute difference X-Y: This would be the measurement of an actual state without inference to the past. I would take this one because it seems more safe.

But would I be able to extract better information from the relative difference in this case?

Vera Maria
  • 31

1 Answers1

1

Yes the relative difference is more informative, but I think you should report both types of analysis as they show different things. Also for the overall relative difference, I would calculate the confidence interval of the final result to determine whether it is statistically significant.

Robert Jones
  • 598